[infinispan-issues] [JBoss JIRA] (ISPN-1942) State transfer timeout confusion

[Dan Berindei (JIRA)]

     [ https://issues.jboss.org/browse/ISPN-1942?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Dan Berindei updated ISPN-1942:
-------------------------------

              Status: Pull Request Sent  (was: Open)
    Git Pull Request: https://github.com/infinispan/infinispan/pull/1035


I fixed the configuration translation to the legacy classes to always set rehashRpcTimeout and I added some INFO log messages for users who configure rehashRpcTimeout/rehashWait.
The state transfer code can keep using rehashRpcTimeout internally until we switch to the new configuration everywhere.

Initially I wanted to implement the stateTransfer/stateRetrieval exclusion as well, but all the stateRetrieval methods actually use stateTransfer under the covers. That made it a little too complicated to keep track of when stateTransfer has been actually "activated" by the user.

There was already a log message warning the users that stateRetrieval is deprecated, we could make it WARN instead of INFO if you think it would help.
                
> State transfer timeout confusion
> --------------------------------
>
>                 Key: ISPN-1942
>                 URL: https://issues.jboss.org/browse/ISPN-1942
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Configuration
>    Affects Versions: 5.1.3.CR1
>            Reporter: Galder Zamarreño
>            Assignee: Dan Berindei
>             Fix For: 5.1.3.FINAL
>
>
> Several things are not right in the forum post above:
> 1. The XML file should not allow both stateRetrieval (deprecated) and stateTransfer to be defined:
> {code}<?xml version="1.0" encoding="UTF-8"?>
> <infinispan
>       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
>       xsi:schemaLocation="urn:infinispan:config:5.1 http://www.infinispan.org/schemas/infinispan-config-5.1.xsd"
>       xmlns="urn:infinispan:config:5.1">
>  
>    <global>
>         <transport clusterName="hotrod-test" distributedSyncTimeout="16000000">
>                 <properties>
>                         <property name="configurationFile" value="/opt/infinispan-5.1.1.FINAL/etc/gossip-router-config.xml"/>
>                 </properties>
>         </transport>
>       <globalJmxStatistics enabled="true"/>
>    </global>
>  
>    <default>
>       <jmxStatistics enabled="true"/>
>       <locking
>          lockAcquisitionTimeout="16000000"
>       />
>  
>       <clustering mode="replicated">
>          <sync replTimeout="16000000"/>
>  
>          <!-- for replication -->
>          <stateRetrieval timeout="16000000"/>
>          <stateTransfer timeout="16000000"/>
>       </clustering>
>    </default>
>  
>    <namedCache name="A"/>
>    <namedCache name="B"/>
>    <namedCache name="C"/>
>    <namedCache name="D"/>
> </infinispan>{code}
> 2. In spite of setting the stateTransfer timeout, state transfer is looking for hash.rehashRpcTimeout which by default is 10 minutes? @Dan, shouldn't the state transfer timeout be passed onto the hash.rehashTimeout? 
> Also @Dan, do we have any docu on how this really should be configured in 5.1?

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira