[JBoss JIRA] (ISPN-10683) Remove CustomFailurePolicy transaction parameters
by Dan Berindei (Jira)
Dan Berindei created ISPN-10683:
-----------------------------------
Summary: Remove CustomFailurePolicy transaction parameters
Key: ISPN-10683
URL: https://issues.jboss.org/browse/ISPN-10683
Project: Infinispan
Issue Type: Task
Components: Core
Affects Versions: 10.0.0.CR2
Reporter: Dan Berindei
Fix For: 10.0.0.Final
The {{CustomFailurePolicy}} javadoc says it should be thread-safe, but the {{javax.transaction.Transaction}} javadoc doesn't say anything about thread safety.
We should replace the {{javax.transaction.Transaction}} with a {{GlobalTransaction}}, which we know is thread-safe.
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
5 years
[JBoss JIRA] (ISPN-10682) Investigate if static final trace variable in classes is helpful
by Will Burns (Jira)
Will Burns created ISPN-10682:
---------------------------------
Summary: Investigate if static final trace variable in classes is helpful
Key: ISPN-10682
URL: https://issues.jboss.org/browse/ISPN-10682
Project: Infinispan
Issue Type: Enhancement
Components: Core
Reporter: Will Burns
Quarkus and native compilation doesn't play nicely with our static final boolean trace fields. The issue is either we have to always have this be true to allow logging on the same native image. Or we have to force it to false, in which case you wouldn't be able to enable tracing without recompiling the native image. We probably should go with the former, but if we do we have to make sure we are doing expensive trace operations inside the block for the case when tracing isn't actually enabled.
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
5 years
[JBoss JIRA] (ISPN-10681) Investigate and possibly convert to StampedLock for OffHeap
by Will Burns (Jira)
[ https://issues.jboss.org/browse/ISPN-10681?page=com.atlassian.jira.plugin... ]
Will Burns updated ISPN-10681:
------------------------------
Description:
We currently use an array of ReentrantReadWriteLock entries to protect critical sections of a off heap map. We shouldn't require reentrancy for our various operations and thus we should be able to hopefully drop in replace these with the equivalent StampedLock write lock instances.
Doug Lea mentions this is the fastest lock they have https://concurrency.markmail.org/thread/kfdv6tg3tnutl4ts
was:We currently use an array of ReentrantReadWriteLock entries to protect critical sections of a off heap map. We shouldn't require reentrancy for our various operations and thus we should be able to hopefully drop in replace these with the equivalent StampedLock write lock instances.
> Investigate and possibly convert to StampedLock for OffHeap
> -----------------------------------------------------------
>
> Key: ISPN-10681
> URL: https://issues.jboss.org/browse/ISPN-10681
> Project: Infinispan
> Issue Type: Enhancement
> Components: Off Heap
> Reporter: Will Burns
> Assignee: Will Burns
> Priority: Major
>
> We currently use an array of ReentrantReadWriteLock entries to protect critical sections of a off heap map. We shouldn't require reentrancy for our various operations and thus we should be able to hopefully drop in replace these with the equivalent StampedLock write lock instances.
> Doug Lea mentions this is the fastest lock they have https://concurrency.markmail.org/thread/kfdv6tg3tnutl4ts
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
5 years
[JBoss JIRA] (ISPN-10681) Investigate and possibly convert to StampedLock for OffHeap
by Will Burns (Jira)
Will Burns created ISPN-10681:
---------------------------------
Summary: Investigate and possibly convert to StampedLock for OffHeap
Key: ISPN-10681
URL: https://issues.jboss.org/browse/ISPN-10681
Project: Infinispan
Issue Type: Enhancement
Components: Off Heap
Reporter: Will Burns
Assignee: Will Burns
We currently use an array of ReentrantReadWriteLock entries to protect critical sections of a off heap map. We shouldn't require reentrancy for our various operations and thus we should be able to hopefully drop in replace these with the equivalent StampedLock write lock instances.
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
5 years