[infinispan-issues] [JBoss JIRA] (ISPN-6209) 2-node cluster with pessimistic lock: Exceptions reported on remote node even if FAIL_SILENTLY flag is used

[Marek Posolda (JIRA)]

     [ https://issues.jboss.org/browse/ISPN-6209?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Marek Posolda updated ISPN-6209:
--------------------------------
    Description: 
When I look at the Javadoc of FAIL_SILENTLY flag, it can be seen:
{code}
This Flag will not be replicated to remote nodes, but it will still protect the invoker from remote exceptions.
{code}

I expect that exception is not reported on any cluster node when this flag is used. However in some situation with Pessimistic locking and with flags ZERO_LOCK_ACQUISITION_TIMEOUT + FAIL_SILENTLY, the exception can be seen in remote node log if the remote node owns the lock.

I am attaching reproducer, which can be used to simulate issue (It's Java class with main method, so can be used in any project with dependency on infinispan 8.1)

All the functionality works as expected, only the exception in the log is annoying.

  was:
When I look at the Javadoc of FAIL_SILENTLY flag, it can be seen:
{code}
This Flag will not be replicated to remote nodes, but it will still protect the invoker from remote exceptions.
{code}

I expect that exception is not reported on any cluster node when this flag is used. However in some situation with Pessimistic locking and with flags ZERO_LOCK_ACQUISITION_TIMEOUT + FAIL_SILENTLY, the exception can be seen in remote node log if the remote node owns the lock.

I am attaching reproducer, which can be used to simulate issue (It's Java class with main method, so can be used in any project with dependency on infinispan 8.1)



> 2-node cluster with pessimistic lock: Exceptions reported on remote node even if FAIL_SILENTLY flag is used
> -----------------------------------------------------------------------------------------------------------
>
>                 Key: ISPN-6209
>                 URL: https://issues.jboss.org/browse/ISPN-6209
>             Project: Infinispan
>          Issue Type: Bug
>    Affects Versions: 8.1.0.Final
>            Reporter: Marek Posolda
>         Attachments: PessimisticLockClusterReproducer.java
>
>
> When I look at the Javadoc of FAIL_SILENTLY flag, it can be seen:
> {code}
> This Flag will not be replicated to remote nodes, but it will still protect the invoker from remote exceptions.
> {code}
> I expect that exception is not reported on any cluster node when this flag is used. However in some situation with Pessimistic locking and with flags ZERO_LOCK_ACQUISITION_TIMEOUT + FAIL_SILENTLY, the exception can be seen in remote node log if the remote node owns the lock.
> I am attaching reproducer, which can be used to simulate issue (It's Java class with main method, so can be used in any project with dependency on infinispan 8.1)
> All the functionality works as expected, only the exception in the log is annoying.



--
This message was sent by Atlassian JIRA
(v6.4.11#64026)