[
https://issues.apache.org/jira/browse/HBASE-15136?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15331148#comment-15331148
]
stack commented on HBASE-15136:
-------------------------------
[~mantonov] Why ain't we working on making this our default? It has FIFO as
behavior when it is not overloaded so we can do stuff like our fastpath speedup
HBASE-16023 when all is 'normal' but then when we start to backup, it moves to
LIFO w/ the controlled delay with shedding, an alogrithim that is pretty basic
and seems to used with some success in a few places at your shop. What you
think? We'd have to do some work to make it so it was as fricitonless as
possible in the 'normal' case..... We can make new issue to make it default?
> Explore different queuing behaviors while busy
> ----------------------------------------------
>
> Key: HBASE-15136
> URL: https://issues.apache.org/jira/browse/HBASE-15136
> Project: HBase
> Issue Type: New Feature
> Components: IPC/RPC
> Reporter: Elliott Clark
> Assignee: Mikhail Antonov
> Priority: Critical
> Fix For: 2.0.0, 1.3.0
>
> Attachments: HBASE-15136-1.2.v1.patch, HBASE-15136-v2.patch,
> deadline_scheduler_v_0_2.patch
>
>
> http://queue.acm.org/detail.cfm?id=2839461
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
