[JBoss JIRA] Created: (ISPN-1414) Introduce a builder than can create an Infinispan CacheManager
by Pete Muir (JIRA)
Introduce a builder than can create an Infinispan CacheManager
--------------------------------------------------------------
Key: ISPN-1414
URL: https://issues.jboss.org/browse/ISPN-1414
Project: Infinispan
Issue Type: Feature Request
Reporter: Pete Muir
Assignee: Manik Surtani
Provide a builder class that allows passing in of SPI instances, rather than doing lookup. This allows for much easier integration, as context from the app server bootstrap can be easily passed in.
For example:
{code}
public class CacheManagerBuilder {
void setClassLoader(ClassLoader cl);
void setTransport(Transport t);
...
Cache start();
}
{code}
We would still want to offer a wrapper around this for Java SE users that would expose a create a CacheManager easily. This factory would want to use a CacheManagerBuilder internally to create the cache, but offer a simplified API. You would want to expose the Builder from the factory as well, offering a sensible set of defaults for SE.
This would break backwards compatibility, as doing new EmbeddedCacheManager etc. would no longer be possible.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months
[JBoss JIRA] (ISPN-1437) Servers need typed configuration and injection of components
by Galder Zamarreño (Created) (JIRA)
Servers need typed configuration and injection of components
------------------------------------------------------------
Key: ISPN-1437
URL: https://issues.jboss.org/browse/ISPN-1437
Project: Infinispan
Issue Type: Enhancement
Components: Configuration
Reporter: Galder Zamarreño
Assignee: Galder Zamarreño
Fix For: 6.0.0.FINAL
Currently Infinispan servers cannot be plugged with executor services for the Netty master and worker threads, which makes them hard to manage in an app server environment where the app server might want to provide its own executors.
This might also require re-thinking how Infinispan servers are configured which currently rely on a bunch of properties. A typed approach would be welcome here.
In Lisbon, we had an interesting discussion around this but in the context of Infinispan core itself, how this is configured, how it has components injected that might be managed by an app server. He has some ideas based on the work he did for Weld, so maybe these can be applied to Infinispan servers too.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months
[JBoss JIRA] (ISPN-61) Transaction 2-phase protocol optimizations
by Galder Zamarreño (Updated) (JIRA)
[ https://issues.jboss.org/browse/ISPN-61?page=com.atlassian.jira.plugin.sy... ]
Galder Zamarreño updated ISPN-61:
---------------------------------
Fix Version/s: 5.1.0.BETA2
(was: 5.1.0.BETA1)
> Transaction 2-phase protocol optimizations
> ------------------------------------------
>
> Key: ISPN-61
> URL: https://issues.jboss.org/browse/ISPN-61
> Project: Infinispan
> Issue Type: Feature Request
> Affects Versions: 4.0.0.ALPHA4
> Reporter: Mircea Markus
> Assignee: Mircea Markus
> Priority: Critical
> Fix For: 5.1.0.BETA2, 5.1.0.FINAL
>
>
> Following 2 optimizations might be implemented with respect to transactions.
> 1) From an email on infinispan-dev (horizon-dev actually):
> if there are only two members int the cluster always use an 1PC. If the 1st phase fails remotely, then also rollback locally. This would reduce one network roundtrip.
> While this is an interesting thought, it does raise the potential for race conditions - since this decision will have to be taken in the TxInterceptor in the beforeCompletion phase of a transaction, and by the time the call gets to the interceptor for replication, the topology may have changed such that you need to replicate to 2 instead of 1 other peer. Which would mean a 2PC again. So it does need some thought.
> 2) when asked to prepare, a participant might return a value indicating that no changes were made (read-only participant), so this one won't need an commit message, so less roundtrip.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months