[JBoss JIRA] (ISPN-6009) Stop using thread-locals for marshalling
by Dan Berindei (JIRA)
[ https://issues.jboss.org/browse/ISPN-6009?page=com.atlassian.jira.plugin.... ]
Dan Berindei reassigned ISPN-6009:
----------------------------------
Assignee: (was: Dan Berindei)
> Stop using thread-locals for marshalling
> ----------------------------------------
>
> Key: ISPN-6009
> URL: https://issues.jboss.org/browse/ISPN-6009
> Project: Infinispan
> Issue Type: Task
> Components: Core, Marshalling
> Affects Versions: 8.0.2.Final
> Reporter: Dan Berindei
> Fix For: 9.0.0.Final
>
>
> Marshalling currently uses a thread-local cache for {{RiverMarshaller}} instances. One of the reasons to cache these {{RiverMarshaller}}s was that they always allocate 2 {{IdentityIntMap}}s, one for the instance cache and one for the class cache.
> But that also creates a problem, because these caches never shrink. If a {{StateResponseCommand}} includes a lot of {{Serializable}} objects, the caches will grow, and then they will never be used untile the next state transfer.
> We should change our internal marshalling API so that it's easy to reuse marshallers without using thread-locals, and to use one-off marshallers for state transfer.
> Currently unmarshalling also requires a thread-local to use the proper classloader. A session-based marshalling API would remove the need for the thread-local (especially important with sequential interceptors).
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
7 years, 11 months
[JBoss JIRA] (ISPN-6009) Stop using thread-locals for marshalling
by Dan Berindei (JIRA)
[ https://issues.jboss.org/browse/ISPN-6009?page=com.atlassian.jira.plugin.... ]
Dan Berindei resolved ISPN-6009.
--------------------------------
Fix Version/s: 9.0.0.Final
Resolution: Done
Fixed with ISPN-6906
> Stop using thread-locals for marshalling
> ----------------------------------------
>
> Key: ISPN-6009
> URL: https://issues.jboss.org/browse/ISPN-6009
> Project: Infinispan
> Issue Type: Task
> Components: Core, Marshalling
> Affects Versions: 8.0.2.Final
> Reporter: Dan Berindei
> Assignee: Dan Berindei
> Fix For: 9.0.0.Final
>
>
> Marshalling currently uses a thread-local cache for {{RiverMarshaller}} instances. One of the reasons to cache these {{RiverMarshaller}}s was that they always allocate 2 {{IdentityIntMap}}s, one for the instance cache and one for the class cache.
> But that also creates a problem, because these caches never shrink. If a {{StateResponseCommand}} includes a lot of {{Serializable}} objects, the caches will grow, and then they will never be used untile the next state transfer.
> We should change our internal marshalling API so that it's easy to reuse marshallers without using thread-locals, and to use one-off marshallers for state transfer.
> Currently unmarshalling also requires a thread-local to use the proper classloader. A session-based marshalling API would remove the need for the thread-local (especially important with sequential interceptors).
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
7 years, 11 months
[JBoss JIRA] (ISPN-9083) Remove support for 'indexed_by_default' option in protobuf schemas
by Adrian Nistor (JIRA)
[ https://issues.jboss.org/browse/ISPN-9083?page=com.atlassian.jira.plugin.... ]
Adrian Nistor updated ISPN-9083:
--------------------------------
Description:
All the behaviour around this option must be removed and the option must be documented as no longer supported in Infinispan 10.
Usage of this option must trigger a warning. Should be entirely removed in 10.1 or some later version right after 10.
was:
All the behaviour around it must be removed and the options must be documented as no longer supported.
Usage of this option must trigger a warning.
> Remove support for 'indexed_by_default' option in protobuf schemas
> ------------------------------------------------------------------
>
> Key: ISPN-9083
> URL: https://issues.jboss.org/browse/ISPN-9083
> Project: Infinispan
> Issue Type: Task
> Reporter: Adrian Nistor
> Assignee: Adrian Nistor
> Fix For: 10.0.0.Final
>
>
> All the behaviour around this option must be removed and the option must be documented as no longer supported in Infinispan 10.
> Usage of this option must trigger a warning. Should be entirely removed in 10.1 or some later version right after 10.
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
7 years, 11 months
[JBoss JIRA] (ISPN-9083) Remove support for 'indexed_by_default' option in protobuf schemas
by Adrian Nistor (JIRA)
[ https://issues.jboss.org/browse/ISPN-9083?page=com.atlassian.jira.plugin.... ]
Adrian Nistor updated ISPN-9083:
--------------------------------
Description:
All the behaviour around this option must be removed and the option must be documented as no longer supported in Infinispan 10.
Usage of this option must trigger a warning. Should be entirely removed in 10.1 or some later version right after 10 - usages will be ignored after this point.
was:
All the behaviour around this option must be removed and the option must be documented as no longer supported in Infinispan 10.
Usage of this option must trigger a warning. Should be entirely removed in 10.1 or some later version right after 10.
> Remove support for 'indexed_by_default' option in protobuf schemas
> ------------------------------------------------------------------
>
> Key: ISPN-9083
> URL: https://issues.jboss.org/browse/ISPN-9083
> Project: Infinispan
> Issue Type: Task
> Reporter: Adrian Nistor
> Assignee: Adrian Nistor
> Fix For: 10.0.0.Final
>
>
> All the behaviour around this option must be removed and the option must be documented as no longer supported in Infinispan 10.
> Usage of this option must trigger a warning. Should be entirely removed in 10.1 or some later version right after 10 - usages will be ignored after this point.
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
7 years, 11 months
[JBoss JIRA] (ISPN-9063) Average stats should be calculated in Nanoseconds but exposed in Milliseconds
by Ryan Emerson (JIRA)
[ https://issues.jboss.org/browse/ISPN-9063?page=com.atlassian.jira.plugin.... ]
Ryan Emerson updated ISPN-9063:
-------------------------------
Sprint: Sprint 9.3.0.Alpha1
> Average stats should be calculated in Nanoseconds but exposed in Milliseconds
> -----------------------------------------------------------------------------
>
> Key: ISPN-9063
> URL: https://issues.jboss.org/browse/ISPN-9063
> Project: Infinispan
> Issue Type: Bug
> Components: Console, JMX, reporting and management
> Affects Versions: 9.2.1.Final
> Reporter: Ryan Emerson
> Assignee: Ryan Emerson
> Fix For: 9.2.2.Final, 9.3.0.Alpha1
>
>
> ISPN-7768 changed the precision of the exposed average-<read|write>time stats from Milliseconds to Nanoseconds. This causes backwards compatibility issues with previous versions of Infinispan and JDG. Therefore we need to ensure that these stats are exposed to the user in Milliseconds, however they should still be calculated in Nanoseconds internally to improve accuracy and avoid rounding errors.
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
7 years, 11 months