[
https://issues.apache.org/jira/browse/HBASE-15137?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mikhail Antonov updated HBASE-15137:
------------------------------------
Attachment: HBASE-15137.v3.patch
Added handling for CQTBE and relevant checks for RetriesWithDetailsExceptions,
which is tricky to unwrap, added a test.
Added a flag to PFFE class indication whether or not operations is guaranteed
to not be attempted by server (for example, CQTBE does guarantee that, but
CallTimeoutException does not)
> CallTimeoutException and CallQueueTooBigException should trigger PFFE
> ---------------------------------------------------------------------
>
> Key: HBASE-15137
> URL: https://issues.apache.org/jira/browse/HBASE-15137
> Project: HBase
> Issue Type: Bug
> Reporter: Elliott Clark
> Assignee: Mikhail Antonov
> Attachments: HBASE-15137.mantonov.patch, HBASE-15137.patch,
> HBASE-15137.v3.patch
>
>
> If a region server is backed up enough that lots of calls are timing out then
> we should think about treating a server as failing.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
