[
https://issues.apache.org/jira/browse/HBASE-14431?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14745379#comment-14745379
]
Samir Ahmic commented on HBASE-14431:
-------------------------------------
bg. Is it a better choice to override hashCode method in AsyncRpcChannel ?
I don't see that method. Can you elaborate ?
BTW is there a reason why
[HBASE-13337|https://issues.apache.org/jira/browse/HBASE-13337] is not
committed to master branch? Without it any testing where restart of servers is
included will cause issues and only in master branch AsyncRpcClient is default
client implementation.
> AsyncRpcClient#removeConnection() never removes connection from connections
> pool if server fails
> ------------------------------------------------------------------------------------------------
>
> Key: HBASE-14431
> URL: https://issues.apache.org/jira/browse/HBASE-14431
> Project: HBase
> Issue Type: Bug
> Components: IPC/RPC
> Affects Versions: 2.0.0, 1.0.2, 1.1.2
> Reporter: Samir Ahmic
> Assignee: Samir Ahmic
> Priority: Critical
>
> I was playing with master branch in distributed mode (3 rs + master +
> backup_master) and notice strange behavior when i was testing this sequence
> of events on single rs: /kill/start/run_balancer while client was writing
> data to cluster (LoadTestTool).
> I have notice that LTT fails with following:
> {code}
> 2015-09-09 11:05:58,364 INFO [main] client.AsyncProcess: #2, waiting for
> some tasks to finish. Expected max=0, tasksInProgress=35
> Exception in thread "main"
> org.apache.hadoop.hbase.client.RetriesExhaustedWithDetailsException: Failed 1
> action: BindException: 1 time,
> at
> org.apache.hadoop.hbase.client.AsyncProcess$BatchErrors.makeException(AsyncProcess.java:228)
> at
> org.apache.hadoop.hbase.client.AsyncProcess$BatchErrors.access$1800(AsyncProcess.java:208)
> at
> org.apache.hadoop.hbase.client.AsyncProcess.waitForAllPreviousOpsAndReset(AsyncProcess.java:1697)
> at
> org.apache.hadoop.hbase.client.BufferedMutatorImpl.backgroundFlushCommits(BufferedMutatorImpl.java:211)
> {code}
> After some digging and adding some more logging in code i have notice that
> following condition in {code}AsyncRpcClient.removeConnection(AsyncRpcChannel
> connection) {code} is never true:
> {code}
> if (connectionInPool == connection) {
> {code}
> causing that {code}AsyncRpcChannel{code} connection is never removed from
> {code}connections{code} pool in case rs fails.
> After changing this condition to:
> {code}
> if (connectionInPool.address.equals(connection.address)) {
> {code}
> issue was resolved and client was removing failed server from connections
> pool.
> I will attach patch after running some more tests.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
