<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
Il 07/11/11 15:05, Galder Zamarreño ha scritto:
<blockquote
cite="mid:4ECE05F1-A90F-4453-8BFF-DE622D8FB1E3@redhat.com"
type="cite">
<pre wrap="">Thx Sebastiano. Would you mind issuing a pull request for your tpcc branch so that we can integrate it into Radargun?</pre>
</blockquote>
Yes of course.<span class="Apple-style-span" style="color: rgb(0, 0,
0); font-family: arial, verdana, sans-serif; font-size: 12px;
font-style: normal; font-variant: normal; font-weight: normal;
letter-spacing: normal; line-height: normal; orphans: 2;
text-align: -webkit-auto; text-indent: 0px; text-transform: none;
white-space: normal; widows: 2; word-spacing: 0px;
-webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0px;
background-color: rgb(240, 240, 248); display: inline !important;
float: none; "><br>
</span>
<blockquote
cite="mid:4ECE05F1-A90F-4453-8BFF-DE622D8FB1E3@redhat.com"
type="cite">
<pre wrap="">
On Nov 7, 2011, at 1:14 PM, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Galder,
thank you for the review. My answers are below inline.
Il 04/11/11 11:02, Galder Zamarreño ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I had the chance to look at the TPCC radargun plug-in you've built (<a class="moz-txt-link-freetext" href="https://github.com/sebastianopeluso/radargun.git">https://github.com/sebastianopeluso/radargun.git</a> - branch TPCC) and I had some comments to make:
1. Instead of using divisions for converting time units, it'd be better to use TimeUnit conversions to make the code more readable.
</pre>
</blockquote>
<pre wrap="">Ok. I will adopt TimeUnit.
</pre>
<blockquote type="cite">
<pre wrap="">2. To make the test run faster, it'd be interesting to transform Order...etc to use Infinispan Externalizers but that's not portable to other frameworks. So, implementing Externalizable could be a good middle ground.
</pre>
</blockquote>
<pre wrap="">Thank you for the suggestion. I will implement Externalizable interface
for the TPC-C domain objects.
</pre>
<blockquote type="cite">
<pre wrap="">3. Is TpccPopulationStage executed in each slave? Or in single slave? Seems like it's only executed in one of the slaves? If so that's fine cos it'd replicate to all right?
</pre>
</blockquote>
<pre wrap="">This is partially true. Each slave populates a slice of the data
container in a such a way that, for each pair of distinct slaves s1, s2,
if s1 populates the slice c1 and s2 populates the slice c2, then the
intersection between c1 and c2 is empty.
In this way we try to parallelize as much as possible cache loading, we
don't perform population in a transactional context and if the cache is
fully replicated then at the end of the population stage each node will
store all loaded data.
</pre>
<blockquote type="cite">
<pre wrap="">4. Has this TPCC benchmark code been run with Infinispan 4?
</pre>
</blockquote>
<pre wrap="">I ran it also on top of Infinispan 5.
</pre>
<blockquote type="cite">
<pre wrap="">Overall looks good from a RadarGun integration point of view, but I'm not familiar with the specifics of TPCC, so someone else should comment on how well the code adheres to the spec.
I've noticed that RadarGun is not up to date with the latest 4.2 Infinispan version and does not have a plugin for Infinispan 5, so I'll update these two pieces.
I plan to only create one plugin for Infinispan 5, which will contain latest Infinispan 5.1 beta.
Cheers,
</pre>
</blockquote>
<pre wrap="">
Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">On Nov 1, 2011, at 4:47 PM, Mircea Markus wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 14 October 2011 10:33:36 GMT+01:00
To: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Cc: Paolo Romano<a class="moz-txt-link-rfc2396E" href="mailto:romano@inesc-id.pt"><romano@inesc-id.pt></a>, Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:msurtani@redhat.com"><msurtani@redhat.com></a>
Hi Mircea,
OK, don't worry! :-)
Cheers
Sebastiano
Il 13/10/11 19:31, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano, right now I'm busy remaking locking/transaction support in 5.1. This should keep me busy for the next 3 weeks or so - after that the plan is to look at tpcc, integrate it and produce performance numbers to go with it.
Sorry for not taking this sooner.
Cheers,
Mircea
On 11 Oct 2011, at 09:49, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
have you had the opportunity to take a look at the Radargun's tpcc branch inhttps://github.com/sebastianopeluso/radargun/tree/tpcc ?
Can I issue anyway a pull request in Github?
Thank you.
Cheers
Sebastiano
Il 29/08/11 11:49, Sebastiano Peluso ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've finalized the implementation of TPC-C code (New Order, Payment and Order Status transactions) within Radargun.
You can find the result as Radargun's tpcc branch at this link:
<a class="moz-txt-link-freetext" href="https://github.com/sebastianopeluso/radargun/tree/tpcc">https://github.com/sebastianopeluso/radargun/tree/tpcc</a>
Can you take a look at the code? I wait for your feedback before issuing a pull request in Github.
Thank you.
Cheers,
Sebastiano
Il 12/07/11 11:20, Sebastiano Peluso ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've just forked Radargun's master version and I will start the integration (following your observations/comments in the last mail) in the next week . I hope to finalize the work by the end of July.
When I will have finished I will issue a pull request.
Thank you for your helpfulness.
Sebastiano
Il 11/07/11 13:28, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Thanks Sebastiano. Once you are finished, can you please issue a pull request for Radargun and we can take it from there? Please let me know if there'a anything I can help you with, e.g. github integration etc.
On 8 Jul 2011, at 15:31, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 13 October 2011 19:31:00 GMT+01:00
To: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>
Cc: Paolo Romano<a class="moz-txt-link-rfc2396E" href="mailto:romano@inesc-id.pt"><romano@inesc-id.pt></a>, Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:msurtani@redhat.com"><msurtani@redhat.com></a>
Hi Sebastiano, right now I'm busy remaking locking/transaction support in 5.1. This should keep me busy for the next 3 weeks or so - after that the plan is to look at tpcc, integrate it and produce performance numbers to go with it.
Sorry for not taking this sooner.
Cheers,
Mircea
On 11 Oct 2011, at 09:49, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
have you had the opportunity to take a look at the Radargun's tpcc branch inhttps://github.com/sebastianopeluso/radargun/tree/tpcc ?
Can I issue anyway a pull request in Github?
Thank you.
Cheers
Sebastiano
Il 29/08/11 11:49, Sebastiano Peluso ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've finalized the implementation of TPC-C code (New Order, Payment and Order Status transactions) within Radargun.
You can find the result as Radargun's tpcc branch at this link:
<a class="moz-txt-link-freetext" href="https://github.com/sebastianopeluso/radargun/tree/tpcc">https://github.com/sebastianopeluso/radargun/tree/tpcc</a>
Can you take a look at the code? I wait for your feedback before issuing a pull request in Github.
Thank you.
Cheers,
Sebastiano
Il 12/07/11 11:20, Sebastiano Peluso ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've just forked Radargun's master version and I will start the integration (following your observations/comments in the last mail) in the next week . I hope to finalize the work by the end of July.
When I will have finished I will issue a pull request.
Thank you for your helpfulness.
Sebastiano
Il 11/07/11 13:28, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Thanks Sebastiano. Once you are finished, can you please issue a pull request for Radargun and we can take it from there? Please let me know if there'a anything I can help you with, e.g. github integration etc.
On 8 Jul 2011, at 15:31, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 11 October 2011 09:49:11 GMT+01:00
To: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Cc: Paolo Romano<a class="moz-txt-link-rfc2396E" href="mailto:romano@inesc-id.pt"><romano@inesc-id.pt></a>
Hi Mircea,
have you had the opportunity to take a look at the Radargun's tpcc branch inhttps://github.com/sebastianopeluso/radargun/tree/tpcc ?
Can I issue anyway a pull request in Github?
Thank you.
Cheers
Sebastiano
Il 29/08/11 11:49, Sebastiano Peluso ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've finalized the implementation of TPC-C code (New Order, Payment and Order Status transactions) within Radargun.
You can find the result as Radargun's tpcc branch at this link:
<a class="moz-txt-link-freetext" href="https://github.com/sebastianopeluso/radargun/tree/tpcc">https://github.com/sebastianopeluso/radargun/tree/tpcc</a>
Can you take a look at the code? I wait for your feedback before issuing a pull request in Github.
Thank you.
Cheers,
Sebastiano
Il 12/07/11 11:20, Sebastiano Peluso ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've just forked Radargun's master version and I will start the integration (following your observations/comments in the last mail) in the next week . I hope to finalize the work by the end of July.
When I will have finished I will issue a pull request.
Thank you for your helpfulness.
Sebastiano
Il 11/07/11 13:28, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Thanks Sebastiano. Once you are finished, can you please issue a pull request for Radargun and we can take it from there? Please let me know if there'a anything I can help you with, e.g. github integration etc.
On 8 Jul 2011, at 15:31, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 31 August 2011 17:36:59 GMT+01:00
To: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>
Cc: Paolo Romano<a class="moz-txt-link-rfc2396E" href="mailto:romano@inesc-id.pt"><romano@inesc-id.pt></a>
Thank you Sebastiano, I'll give it a look this/next week.
On 29 Aug 2011, at 11:49, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've finalized the implementation of TPC-C code (New Order, Payment and Order Status transactions) within Radargun.
You can find the result as Radargun's tpcc branch at this link:
<a class="moz-txt-link-freetext" href="https://github.com/sebastianopeluso/radargun/tree/tpcc">https://github.com/sebastianopeluso/radargun/tree/tpcc</a>
Can you take a look at the code? I wait for your feedback before issuing a pull request in Github.
Thank you.
Cheers,
Sebastiano
Il 12/07/11 11:20, Sebastiano Peluso ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've just forked Radargun's master version and I will start the integration (following your observations/comments in the last mail) in the next week . I hope to finalize the work by the end of July.
When I will have finished I will issue a pull request.
Thank you for your helpfulness.
Sebastiano
Il 11/07/11 13:28, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Thanks Sebastiano. Once you are finished, can you please issue a pull request for Radargun and we can take it from there? Please let me know if there'a anything I can help you with, e.g. github integration etc.
On 8 Jul 2011, at 15:31, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 29 August 2011 11:49:17 GMT+01:00
To: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Cc: Paolo Romano<a class="moz-txt-link-rfc2396E" href="mailto:romano@inesc-id.pt"><romano@inesc-id.pt></a>
Hi Mircea,
I've finalized the implementation of TPC-C code (New Order, Payment and Order Status transactions) within Radargun.
You can find the result as Radargun's tpcc branch at this link:
<a class="moz-txt-link-freetext" href="https://github.com/sebastianopeluso/radargun/tree/tpcc">https://github.com/sebastianopeluso/radargun/tree/tpcc</a>
Can you take a look at the code? I wait for your feedback before issuing a pull request in Github.
Thank you.
Cheers,
Sebastiano
Il 12/07/11 11:20, Sebastiano Peluso ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
I've just forked Radargun's master version and I will start the integration (following your observations/comments in the last mail) in the next week . I hope to finalize the work by the end of July.
When I will have finished I will issue a pull request.
Thank you for your helpfulness.
Sebastiano
Il 11/07/11 13:28, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Thanks Sebastiano. Once you are finished, can you please issue a pull request for Radargun and we can take it from there? Please let me know if there'a anything I can help you with, e.g. github integration etc.
On 8 Jul 2011, at 15:31, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 12 July 2011 10:20:33 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Hi Mircea,
I've just forked Radargun's master version and I will start the integration (following your observations/comments in the last mail) in the next week . I hope to finalize the work by the end of July.
When I will have finished I will issue a pull request.
Thank you for your helpfulness.
Sebastiano
Il 11/07/11 13:28, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Thanks Sebastiano. Once you are finished, can you please issue a pull request for Radargun and we can take it from there? Please let me know if there'a anything I can help you with, e.g. github integration etc.
On 8 Jul 2011, at 15:31, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 11 July 2011 13:28:15 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
Thanks Sebastiano. Once you are finished, can you please issue a pull request for Radargun and we can take it from there? Please let me know if there'a anything I can help you with, e.g. github integration etc.
On 8 Jul 2011, at 15:31, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 11 July 2011 13:28:15 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Thanks Sebastiano. Once you are finished, can you please issue a pull request for Radargun and we can take it from there? Please let me know if there'a anything I can help you with, e.g. github integration etc.
On 8 Jul 2011, at 15:31, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 8 July 2011 15:31:03 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Hi Mircea,
thank you for your reply. My answers to your questions/comments are inline.
Il 08/07/11 11:33, Mircea Markus ha scritto:
</pre>
<blockquote type="cite">
<pre wrap="">Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
</pre>
</blockquote>
<pre wrap="">YES.
</pre>
<blockquote type="cite">
<pre wrap="">Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
</pre>
</blockquote>
<pre wrap="">Ok. Your observation is correct. In this way the code is more modular and we can have both the web session and the tpcc benchmarks.
</pre>
<blockquote type="cite">
<pre wrap="">- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes? In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
</pre>
</blockquote>
<pre wrap="">This is partially correct. The "isPrimary" condition is evaluated only if the Primary-Backup replication mechanism is enabled because the code has been written in order to test this replication mechanism too. In this case only the primary node can perform write transactions.
But if 2-Phase-Commit is enabled (lines 72-83 in TPCCTerminal), transactions (read and write) are spread over all nodes. This is always the case, because we don't have Primary-Backup in official Infinispan releases.
</pre>
<blockquote type="cite">
<pre wrap="">- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
</pre>
</blockquote>
<pre wrap="">Sorry, this is a mistake :-) .
</pre>
<blockquote type="cite">
<pre wrap="">- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
</pre>
</blockquote>
<pre wrap="">I agree. I can create one class for each tpcc transaction and assign each "case" block (referring to the current implementation of the executeTransaction method) to the corresponding class.
</pre>
<blockquote type="cite">
<pre wrap="">- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
</pre>
</blockquote>
<pre wrap="">Ok. I'll remove .idea directory.
In addition, I'll insert tpcc implementation in a forked version of radargun's master in order to benefit from merge features provided by git/github.
</pre>
<blockquote type="cite">
<pre wrap="">Cheers,
Mircea
</pre>
</blockquote>
<pre wrap="">Thank you again.
Cheers,
Sebastiano
</pre>
<blockquote type="cite">
<pre wrap="">[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:msurtani@redhat.com"><msurtani@redhat.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 8 July 2011 12:04:56 GMT+01:00
To: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Cc: Dan Berindei<a class="moz-txt-link-rfc2396E" href="mailto:dberinde@redhat.com"><dberinde@redhat.com></a>
On 8 Jul 2011, at 11:41, Mircea Markus wrote:
</pre>
<blockquote type="cite">
<pre wrap="">I did, search this email for "Here are some observations I have after looking".
</pre>
</blockquote>
<pre wrap="">Ok, didn't read the forwarded email. :)
--
Manik Surtani
<a class="moz-txt-link-abbreviated" href="mailto:manik@jboss.org">manik@jboss.org</a>
twitter.com/maniksurtani
Lead, Infinispan
<a class="moz-txt-link-freetext" href="http://www.infinispan.org">http://www.infinispan.org</a>
</pre>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 8 July 2011 11:41:42 GMT+01:00
To: Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:msurtani@redhat.com"><msurtani@redhat.com></a>
Cc: Dan Berindei<a class="moz-txt-link-rfc2396E" href="mailto:dberinde@redhat.com"><dberinde@redhat.com></a>
I did, search this email for "Here are some observations I have after looking".
On 8 Jul 2011, at 11:39, Manik Surtani wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Could you give them some pointers on how to make this easier to integrate?
Sent from my mobile phone
On 8 Jul 2011, at 11:36, Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a> wrote:
</pre>
<blockquote type="cite">
<pre wrap="">FYI, I reviewed the code and is far from being integrable in its current shape. The idea is really nice though, hope they'll update and we can at least take some of it.
Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Date: 8 July 2011 11:33:41 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes?
In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
Cheers,
Mircea
[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:msurtani@redhat.com"><msurtani@redhat.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 8 July 2011 11:39:37 GMT+01:00
To: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Cc: Dan Berindei<a class="moz-txt-link-rfc2396E" href="mailto:dberinde@redhat.com"><dberinde@redhat.com></a>, Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
Could you give them some pointers on how to make this easier to integrate?
Sent from my mobile phone
On 8 Jul 2011, at 11:36, Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a> wrote:
</pre>
<blockquote type="cite">
<pre wrap="">FYI, I reviewed the code and is far from being integrable in its current shape. The idea is really nice though, hope they'll update and we can at least take some of it.
Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Date: 8 July 2011 11:33:41 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes?
In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
Cheers,
Mircea
[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Fwd: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 8 July 2011 11:36:29 GMT+01:00
To: Dan Berindei<a class="moz-txt-link-rfc2396E" href="mailto:dberinde@redhat.com"><dberinde@redhat.com></a>, Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
FYI, I reviewed the code and is far from being integrable in its current shape. The idea is really nice though, hope they'll update and we can at least take some of it.
Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Date: 8 July 2011 11:33:41 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes?
In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
Cheers,
Mircea
[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 8 July 2011 11:33:41 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes?
In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
Cheers,
Mircea
[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 8 July 2011 11:33:41 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Hi Sebastiano,
Very interesting stuff!
I think this is a nice approach of benchmarking transactions with a controlled degree contention. This is a feature currently missing in Radargun, but very important especially considering the stuff you guys are doing and the locking improvements we have scheduled for 5.1 release.
Do you consider contributing this back to Radargun?
Here are some observations I have after looking at the code:
- The code should not be in the PutGetStressor, as this is used for benchmarking a web-session-like access pattern which we don't want to replace. It might make more sense to be placed in a stage of its own, e.g. org.radargun.stages.TpccBenchmarkStage
- "isPrimary" argument of TPCCTerminal.choiceTransaction is true if the node is a jgroups coordinator. This has some limitations:
- other data grid providers than Infinispan might not have an "coordinator" concept, and this might be difficult to infer
- ATM only the coordinator does writes (on multiple threads). IMO this is too restricting as it doesn't induce deadlock situations that might appear when data is written by multiple nodes. Wouldn't it make more sense to spread the writes over all the nodes?
In order to overcome the dependency on jgroup's coordinator, you can use instead the AbstractDistStage.slaveIndex. This is a Radargun-level view of the benchmark cluster, independent of the benchmarked product
- PutGetStressor compares Strings based on the == operator, which is incorrect. e.g. e.getClass().getName() == "org.infinispan.CacheException"
- As an design suggestion, I think it would make sense to have an TpccTransaction interface with an executeTransaction(CacheWrapper cacheWrapper) methods, with specialized implementations: NewOrderTransaction, PayementTransaction, OrderStatusTransaction. Basically to move all the code TPCCTerminal.executeTransaction into these three classes. Such an approach would reduce coupling between the transactions and most importantly would offer a simple way of adding new tpcc transactions to the system, by implementing the TpccTransaction interface.
- class javadoc are the onese generater by IDEA
- adding .idea directory in github might cause troubles for people using IDEA and forking your repo; I think the dir should be removed
Another observation is about the way the project is managed in github: it is not a fork of Infinispan, but it was downloaded and committed as a totally different project. This means that all the nice review and merge feature that github and git offer cannot be used, but the merge needs to be made "by hand", comparing the source trees etc. I think it would make more sense to fork[1] from radargun's master[2] and apply all these nice feature on a fork. What do you think?
Cheers,
Mircea
[1]<a class="moz-txt-link-freetext" href="http://help.github.com/fork-a-repo/">http://help.github.com/fork-a-repo/</a>
[2]<a class="moz-txt-link-freetext" href="https://github.com/radargun/radargun">https://github.com/radargun/radargun</a>
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________">http://p.sf.net/sfu/splunk-d2d-c2_______________________________________________</a>
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 28 June 2011 13:36:40 GMT+01:00
To: Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
<a class="moz-txt-link-abbreviated" href="mailto:Cc:cloudtm-discussion@lists.sourceforge.net">Cc:cloudtm-discussion@lists.sourceforge.net</a>
Didn't look at it in details but looks promising. I'll take a look later on this week.
On 22 Jun 2011, at 18:44, Manik Surtani wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Cool. I'm adding Mircea in cc, who maintains the upstream Radargun. Mircea, does it make sense to have this in Radargun upstream?
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">--
Manik Surtani
<a class="moz-txt-link-abbreviated" href="mailto:manik@jboss.org">manik@jboss.org</a>
twitter.com/maniksurtani
Lead, Infinispan
<a class="moz-txt-link-freetext" href="http://www.infinispan.org">http://www.infinispan.org</a>
</pre>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 28 June 2011 13:36:40 GMT+01:00
To: Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
<a class="moz-txt-link-abbreviated" href="mailto:Cc:cloudtm-discussion@lists.sourceforge.net">Cc:cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Didn't look at it in details but looks promising. I'll take a look later on this week.
On 22 Jun 2011, at 18:44, Manik Surtani wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Cool. I'm adding Mircea in cc, who maintains the upstream Radargun. Mircea, does it make sense to have this in Radargun upstream?
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">--
Manik Surtani
<a class="moz-txt-link-abbreviated" href="mailto:manik@jboss.org">manik@jboss.org</a>
twitter.com/maniksurtani
Lead, Infinispan
<a class="moz-txt-link-freetext" href="http://www.infinispan.org">http://www.infinispan.org</a>
</pre>
</blockquote>
<pre wrap="">------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/splunk-d2d-c2">http://p.sf.net/sfu/splunk-d2d-c2</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 27 June 2011 17:01:01 GMT+01:00
To: Paolo Romano<a class="moz-txt-link-rfc2396E" href="mailto:romano@inesc-id.pt"><romano@inesc-id.pt></a>
Cc: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>,<a class="moz-txt-link-abbreviated" href="mailto:mmarkus@redhat.com">mmarkus@redhat.com</a>, Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
Thanks wiling to contribute this!
I've received my account details and I'll take a look this/next week at it.
Cheers,
Mircea
On 22 Jun 2011, at 19:41, Paolo Romano wrote:
</pre>
<blockquote type="cite">
<pre wrap="">I am afraid that Mircea will not be able to log-in to the wiki.
Sebastiano, please, can you create an account for him?
Cheers,
Paolo
On 6/22/11 6:44 PM, Manik Surtani wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Cool. I'm adding Mircea in cc, who maintains the upstream Radargun. Mircea, does it make sense to have this in Radargun upstream?
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">--
Manik Surtani
<a class="moz-txt-link-abbreviated" href="mailto:manik@jboss.org">manik@jboss.org</a>
twitter.com/maniksurtani
Lead, Infinispan
<a class="moz-txt-link-freetext" href="http://www.infinispan.org">http://www.infinispan.org</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mircea.markus@jboss.com"><mircea.markus@jboss.com></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 27 June 2011 17:01:01 GMT+01:00
To: Paolo Romano<a class="moz-txt-link-rfc2396E" href="mailto:romano@inesc-id.pt"><romano@inesc-id.pt></a>
Cc: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>,<a class="moz-txt-link-abbreviated" href="mailto:mmarkus@redhat.com">mmarkus@redhat.com</a>, Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
Thanks wiling to contribute this!
I've received my account details and I'll take a look this/next week at it.
Cheers,
Mircea
On 22 Jun 2011, at 19:41, Paolo Romano wrote:
</pre>
<blockquote type="cite">
<pre wrap="">I am afraid that Mircea will not be able to log-in to the wiki.
Sebastiano, please, can you create an account for him?
Cheers,
Paolo
On 6/22/11 6:44 PM, Manik Surtani wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Cool. I'm adding Mircea in cc, who maintains the upstream Radargun. Mircea, does it make sense to have this in Radargun upstream?
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">--
Manik Surtani
<a class="moz-txt-link-abbreviated" href="mailto:manik@jboss.org">manik@jboss.org</a>
twitter.com/maniksurtani
Lead, Infinispan
<a class="moz-txt-link-freetext" href="http://www.infinispan.org">http://www.infinispan.org</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Paolo Romano<a class="moz-txt-link-rfc2396E" href="mailto:romano@inesc-id.pt"><romano@inesc-id.pt></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 22 June 2011 19:41:00 GMT+01:00
To: Sebastiano Peluso<a class="moz-txt-link-rfc2396E" href="mailto:peluso@gsd.inesc-id.pt"><peluso@gsd.inesc-id.pt></a>,<a class="moz-txt-link-abbreviated" href="mailto:mmarkus@redhat.com">mmarkus@redhat.com</a>
Cc: Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
I am afraid that Mircea will not be able to log-in to the wiki.
Sebastiano, please, can you create an account for him?
Cheers,
Paolo
On 6/22/11 6:44 PM, Manik Surtani wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Cool. I'm adding Mircea in cc, who maintains the upstream Radargun. Mircea, does it make sense to have this in Radargun upstream?
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">--
Manik Surtani
<a class="moz-txt-link-abbreviated" href="mailto:manik@jboss.org">manik@jboss.org</a>
twitter.com/maniksurtani
Lead, Infinispan
<a class="moz-txt-link-freetext" href="http://www.infinispan.org">http://www.infinispan.org</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 22 June 2011 18:44:50 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
Cc: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mmarkus@redhat.com"><mmarkus@redhat.com></a>
Cool. I'm adding Mircea in cc, who maintains the upstream Radargun. Mircea, does it make sense to have this in Radargun upstream?
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">--
Manik Surtani
<a class="moz-txt-link-abbreviated" href="mailto:manik@jboss.org">manik@jboss.org</a>
twitter.com/maniksurtani
Lead, Infinispan
<a class="moz-txt-link-freetext" href="http://www.infinispan.org">http://www.infinispan.org</a>
</pre>
</blockquote>
<pre wrap="">Begin forwarded message:
</pre>
<blockquote type="cite">
<pre wrap="">From: Manik Surtani<a class="moz-txt-link-rfc2396E" href="mailto:manik@jboss.org"><manik@jboss.org></a>
Subject: Re: [Cloudtm-discussion] RadargunTPCC benchmark
Date: 22 June 2011 18:44:50 GMT+01:00
<a class="moz-txt-link-abbreviated" href="mailto:To:cloudtm-discussion@lists.sourceforge.net">To:cloudtm-discussion@lists.sourceforge.net</a>
Cc: Mircea Markus<a class="moz-txt-link-rfc2396E" href="mailto:mmarkus@redhat.com"><mmarkus@redhat.com></a>
<a class="moz-txt-link-abbreviated" href="mailto:Reply-To:cloudtm-discussion@lists.sourceforge.net">Reply-To:cloudtm-discussion@lists.sourceforge.net</a>
Cool. I'm adding Mircea in cc, who maintains the upstream Radargun. Mircea, does it make sense to have this in Radargun upstream?
On 22 Jun 2011, at 18:06, Sebastiano Peluso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi all,
I have added a new subsection
(<a class="moz-txt-link-freetext" href="https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC">https://cloudtm.ist.utl.pt/wiki/index.php/RadargunTPCC</a>), within the
"Shared software resources" section on the Cloud-TM Wiki, in order to
share information about the RadargunTPCC benchmark.
RadargunTPCC benchmark is an implementation of the Transaction Profiles
defined in the TPC Benchmark C specification [1] within Radargun
Benchamark [2], and it has been designed with the purpose of executing
the TPC-C benchmark against Infinispan.
The source code is available at this link:
<a class="moz-txt-link-freetext" href="https://github.com/cloudtm/RadargunTPCC">https://github.com/cloudtm/RadargunTPCC</a>
Regards,
Sebastiano Peluso
[1]<a class="moz-txt-link-freetext" href="http://www.tpc.org/tpcc/spec/tpcc_current.pdf">http://www.tpc.org/tpcc/spec/tpcc_current.pdf</a>
[2]<a class="moz-txt-link-freetext" href="http://sourceforge.net/apps/trac/radargun/wiki/WikiStart">http://sourceforge.net/apps/trac/radargun/wiki/WikiStart</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
<pre wrap="">--
Manik Surtani
<a class="moz-txt-link-abbreviated" href="mailto:manik@jboss.org">manik@jboss.org</a>
twitter.com/maniksurtani
Lead, Infinispan
<a class="moz-txt-link-freetext" href="http://www.infinispan.org">http://www.infinispan.org</a>
------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today.
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/quest-sfdev2dev">http://p.sf.net/sfu/quest-sfdev2dev</a>
_______________________________________________
Cloudtm-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Cloudtm-discussion@lists.sourceforge.net">Cloudtm-discussion@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion">https://lists.sourceforge.net/lists/listinfo/cloudtm-discussion</a>
</pre>
</blockquote>
</blockquote>
<pre wrap="">--
Galder Zamarreño
Sr. Software Engineer
Infinispan, JBoss Cache
</pre>
</blockquote>
<pre wrap="">
_______________________________________________
infinispan-dev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:infinispan-dev@lists.jboss.org">infinispan-dev@lists.jboss.org</a>
<a class="moz-txt-link-freetext" href="https://lists.jboss.org/mailman/listinfo/infinispan-dev">https://lists.jboss.org/mailman/listinfo/infinispan-dev</a>
</pre>
</blockquote>
<pre wrap="">
--
Galder Zamarreño
Sr. Software Engineer
Infinispan, JBoss Cache
_______________________________________________
infinispan-dev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:infinispan-dev@lists.jboss.org">infinispan-dev@lists.jboss.org</a>
<a class="moz-txt-link-freetext" href="https://lists.jboss.org/mailman/listinfo/infinispan-dev">https://lists.jboss.org/mailman/listinfo/infinispan-dev</a>
</pre>
</blockquote>
<br>
</body>
</html>