[jboss-jira] [JBoss JIRA] Commented: (JGRP-610) Problems with large return value and FC

[Bela Ban (JIRA)]

    [ http://jira.jboss.com/jira/browse/JGRP-610?page=comments#action_12385326 ] 
            
Bela Ban commented on JGRP-610:
-------------------------------

Detailed explanation:
- A unicast is received. This acquires a (sender-based) lockin UNICAST on its way up
- When the message is received, the RpcDispatcher sends the large return value back *on the same thread* (the lock in UNICAST is not released)
- FRAG2 splits the return message into many smaller messages
- All messages are sent down, but when max_credits is exceeded, FC.down() blocks. This means the lock in UNICAST is *still* not released
- On the receiver side (sender == receiver), incoming calls cannot reach FC due to the lock in UNICAST (because FC is somewhere above 
  UNICAST), therefore FC will not send new credits (via OOB) to unblock the sender
- The only way credits will be received is for the sender to periodically (max_block_time ms) send an OOB credit request and unblock when it 
  receives the credit response
- However, even OOB messages will not get processed when the default thread pool in TP has the queue *disabled* and all threads are *busy* !

> Problems with large return value and FC
> ---------------------------------------
>
>                 Key: JGRP-610
>                 URL: http://jira.jboss.com/jira/browse/JGRP-610
>             Project: JGroups
>          Issue Type: Bug
>            Reporter: Bela Ban
>         Assigned To: Bela Ban
>             Fix For: 2.6
>
>
> When sending a return value whose size is larger than FC.max_credits, the call hangs. Unit test RpcDispatcher.testLargeReturnValueUnicastCall() reproduces the issue.

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