[
https://issues.apache.org/jira/browse/HBASE-17852?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16234841#comment-16234841
]
Andrew Purtell commented on HBASE-17852:
----------------------------------------
bq. This is what Phoenix does for indexing AFAIK
This has brought down big clusters where I work. I'm not saying don't do it,
but whatever depends on cross-server RPC should learn from the Phoenix example:
- Never block on those remote RPCs in critical sections, holding locks,
especially if you are running in a RPC handler already
- Don't expect the remote resource to be available
- Fail as quickly as possible and retry/clean up later
Let me just assume this stuff is handled, but a walk through of what happens
when the backup table goes away in different scenarios would be good.
We are also going to have these same issues when we migrate replication
tracking state away from ZooKeeper into a system table. At some point we have
to communicate off server for resilient state tracking, no ways around it.
Maybe we can build up some kind of library for cross server RPC.
> Add Fault tolerance to HBASE-14417 (Support bulk loaded files in incremental
> backup)
> ------------------------------------------------------------------------------------
>
> Key: HBASE-17852
> URL: https://issues.apache.org/jira/browse/HBASE-17852
> Project: HBase
> Issue Type: Sub-task
> Reporter: Vladimir Rodionov
> Assignee: Vladimir Rodionov
> Priority: Major
> Fix For: 2.0.0-beta-1
>
> Attachments: HBASE-17852-v1.patch, HBASE-17852-v2.patch,
> HBASE-17852-v3.patch, HBASE-17852-v4.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
