[
https://issues.apache.org/jira/browse/HBASE-10357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13981776#comment-13981776
]
Enis Soztutar commented on HBASE-10357:
---------------------------------------
- Can we default to 200ms or smt. 1s seems long. However, this rpc request
seems that it would wait for the whole batch to complete. Maybe we should have
different confs for switching from one replica to another. If the scan usually
takes around that timeout, then we can be switching back and forth a lot
causing open+close scanners very frequently.
{code}
+ this.configuration.getInt("hbase.client.primaryCallTimeout.scan",
1000000); // 1000 ms
{code}
- Instead of wrapping ScannerCallable everytime, can we check the consistency
and not wrap it if STRONG?
- in addCallsForCurrentReplica, the location is always from the primary
replica?
- Would be good to have a test where we switch in the middle of the scan as
well. I think current test only switching at start.
> Failover RPC's for scans
> ------------------------
>
> Key: HBASE-10357
> URL: https://issues.apache.org/jira/browse/HBASE-10357
> Project: HBase
> Issue Type: Sub-task
> Components: Client
> Reporter: Enis Soztutar
> Fix For: 0.99.0
>
> Attachments: 10357-1.txt, 10357-2.txt, 10357-3.2.txt, 10357-3.txt
>
>
> This is extension of HBASE-10355 to add failover support for scans.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
