[
https://issues.apache.org/jira/browse/HBASE-3787?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13678414#comment-13678414
]
Sergey Shelukhin commented on HBASE-3787:
-----------------------------------------
MRU XID list in the paper actually looks somewhat similar to incrementing
nonces above. We don't have the problem of clients with single vs. multiple
threads etc. because we control the client, incrementing nonces would be per
region per client to roll forward easier.
I'd say response cache may not be feasible if response contains KVs, that's too
much to cache imho, especially given the nature of distributed retries (they
can take longer time between retries).
Main problem here as far as I see it is doing things between servers. For that,
somewhat complicated things will have to happen. I don't see design consensus
here... if we think this is important, after HBaseCon, based on current patch,
I can implement one of the complex schemes to see what it looks like. What do
you think it should be? Incrementing nonces + "nonce WAL" as described above
sound good?
> Increment is non-idempotent but client retries RPC
> --------------------------------------------------
>
> Key: HBASE-3787
> URL: https://issues.apache.org/jira/browse/HBASE-3787
> Project: HBase
> Issue Type: Bug
> Components: Client
> Affects Versions: 0.94.4, 0.95.2
> Reporter: dhruba borthakur
> Assignee: Sergey Shelukhin
> Priority: Critical
> Fix For: 0.95.1
>
> Attachments: HBASE-3787-partial.patch, HBASE-3787-v0.patch,
> HBASE-3787-v1.patch, HBASE-3787-v2.patch, HBASE-3787-v3.patch,
> HBASE-3787-v4.patch, HBASE-3787-v5.patch, HBASE-3787-v5.patch
>
>
> The HTable.increment() operation is non-idempotent. The client retries the
> increment RPC a few times (as specified by configuration) before throwing an
> error to the application. This makes it possible that the same increment call
> be applied twice at the server.
> For increment operations, is it better to use
> HConnectionManager.getRegionServerWithoutRetries()? Another option would be
> to enhance the IPC module to make the RPC server correctly identify if the
> RPC is a retry attempt and handle accordingly.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
