[jboss-jira] [JBoss JIRA] Updated: (JBCLUSTER-223) Support for distributed locking

[Brian Stansberry (JIRA)]

     [ https://jira.jboss.org/jira/browse/JBCLUSTER-223?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Brian Stansberry updated JBCLUSTER-223:
---------------------------------------

    Description: 
Useful for clustered DeploymentRepository and for ensuring only one node accesses a web session or SFSB at a time.

This actually exists in JBoss AS cluster module, so this JIRA is largely about extracting it out of the AS codebase.

This is different from the JGroups DistributedLockManager isn two main respects:

1) Built on HAPartition instead of RpcDispatcher.  This is actually a negative, but I want this to be able to use the same channel the AS's HAPartition uses.  Perhaps we'll develop a separate API and an adapter implementation that directly uses RpcDispatcher to provide flexibility going forward (see JBAS-7850).

2) Acquiring a lock requires a callout to a pluggable LocalLockHandler. This is to support more sophisticated interaction with the service using the lock manager.

It doesn't use the two-phase approach the JGroups DistributedLockManager adopts either.

  was:
Useful for clustered DeploymentRepository and for ensuring only one node accesses a web session or SFSB at a time.

This will be different from the JGroups DistributedLockManager is two main respects:

1) Built on HAPartition instead of RpcDispatcher.  This is actually a negative, but I want this to be able to use the same channel the AS's HAPartition uses.  Perhaps I'll develop a separate API and an adapter implementation that directly uses RpcDispatcher to provide flexibility going forward.

2) Acquiring a lock requires a callout to a pluggable LocalLockHandler. This is to support more sophisticated interaction with the service using the lock manager.

I probably won't use the two-phase approach the JGroups DistributedLockManager adopts either.

       Assignee:     (was: Brian Stansberry)


Unassigning from myself so others working on the Infinispan migration feel free to pick it up.

> Support for distributed locking
> -------------------------------
>
>                 Key: JBCLUSTER-223
>                 URL: https://jira.jboss.org/jira/browse/JBCLUSTER-223
>             Project: JBoss Clustering
>          Issue Type: Feature Request
>      Security Level: Public(Everyone can see) 
>          Components: HA-Server-API
>            Reporter: Brian Stansberry
>             Fix For: HA-Server-API 2.0.0.Beta1
>
>
> Useful for clustered DeploymentRepository and for ensuring only one node accesses a web session or SFSB at a time.
> This actually exists in JBoss AS cluster module, so this JIRA is largely about extracting it out of the AS codebase.
> This is different from the JGroups DistributedLockManager isn two main respects:
> 1) Built on HAPartition instead of RpcDispatcher.  This is actually a negative, but I want this to be able to use the same channel the AS's HAPartition uses.  Perhaps we'll develop a separate API and an adapter implementation that directly uses RpcDispatcher to provide flexibility going forward (see JBAS-7850).
> 2) Acquiring a lock requires a callout to a pluggable LocalLockHandler. This is to support more sophisticated interaction with the service using the lock manager.
> It doesn't use the two-phase approach the JGroups DistributedLockManager adopts either.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira