[
https://issues.apache.org/jira/browse/HBASE-13919?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14708254#comment-14708254
]
Lars Hofhansl commented on HBASE-13919:
---------------------------------------
bq. different operations (put, get or multi-get, coprocessor call, etc) respect
different parameters for RPC timeout, operation timeout and retries depending
on the code path
Right. And that, IMHO, does not make any sense. We have many options that do
the right thing, and depending on how you use HBase you need to set one or the
other, or both, etc. I just meant I'll document them all first (here in the
jira) and then I'll make a proposal about which ones to remove, which to
consolidate, etc.
> Rationalize Client Timeout
> --------------------------
>
> Key: HBASE-13919
> URL: https://issues.apache.org/jira/browse/HBASE-13919
> Project: HBase
> Issue Type: Brainstorming
> Reporter: Lars Hofhansl
>
> There are current many setting that influence how/when an HBase client times
> out.
> This is hard to configure, hard to understand, and badly documented.
> Instead there should as few tweaks as needed, and their setting should make
> sense.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
