[infinispan-dev] introducing an artificial network delay

Sanne Grinovero sanne.grinovero at gmail.com
Wed Jul 7 13:57:27 EDT 2010


2010/7/6 Mircea Markus <mircea.markus at jboss.com>:
>
> On 6 Jul 2010, at 16:06, Manik Surtani wrote:
>
>> I have used this technique in some other tests as well.  Have a look at RehashAfterPartitionMergeTest, for example, on line 39:
>>
>>       http://viewvc.jboss.org/cgi-bin/viewvc.cgi/infinispan/trunk/core/src/test/java/org/infinispan/distribution/rehash/RehashAfterPartitionMergeTest.java?view=markup
>>
>> Perhaps it makes sense to wrap this stuff up in a helper method on TestingUtil so it can be reused easily, perhaps something like:
> OOT - About TestingUtil - all these methods are generally used in classes that extend AbstractCacheTest (MultipleCacheManagersTest/SingleCacheManagerTest extend it as well). Wouldn't it be more API friendly to have these methods on the AbstractCacheTest rather than on TestingUtil?

My use case is not extending AbstractCacheTest, so I'll start
implementing it as static helper, I promise I'll add a nice delegate
from AbstractCacheTest as soon as you get a usecase.

>>
>>       public DISCARD addDISCARD(Cache c) { ... }
>>
>> and
>>
>>       public DELAY addDELAY(Cache c) { ... } ?
>>
>> We'd probably want the reverse as well:
>>
>>       public void removeDISCARD(Cache c) { ... }
>>       public void removeDELAY(Cache c) { ... }
>>

I've opened ISPN-531 and plan to contribute what you suggested.

cheers,
Sanne

>>
>>
>>
>>
>> On 6 Jul 2010, at 13:54, Bela Ban wrote:
>>
>>>
>>>
>>> Sanne Grinovero wrote:
>>>> Hello,
>>>> I'm running some performance tests; my goal is to alter the code for
>>>> the Lucene Directory to use optimal batching and Flags, but to make
>>>> sure I'm improving performance for DIST and REPL configurations the
>>>> only option I'm having is to run it on a real cluster, which is not
>>>> doable right now.
>>>>
>>>> So as I'm running tests on my laptop, network IO is not real and I
>>>> suppose it's not representative of real life bottlenecks.
>>>>
>>>> Is there some option I could set in the jgroups configuration to
>>>> introduce an arbitrary delay,
>>>
>>> Yes, DELAY ! :-)
>>>
>>> Options in_delay and out_delay add some ms to every incoming or outgoing
>>> message. Take a look at the code, DELAY is only ~ 80 lines, so you could
>>> also write your own, which introduces varying delays.
>>>
>>>> or even some degree of random packet losses?
>>>
>>> Yes, DISCARD ! Use 'up' and 'down' to introduce random discards, e.g.
>>> DISCARD up=".4" means that 40% of incoming messages are discarded.
>>>
>>>
>>> --
>>> Bela Ban
>>> Lead JGroups / Clustering Team
>>> JBoss
>>> _______________________________________________
>>> infinispan-dev mailing list
>>> infinispan-dev at lists.jboss.org
>>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>>
>> --
>> Manik Surtani
>> manik at jboss.org
>> Lead, Infinispan
>> Lead, JBoss Cache
>> http://www.infinispan.org
>> http://www.jbosscache.org
>>
>>
>>
>>
>>
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>
>
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>



More information about the infinispan-dev mailing list