[
https://issues.apache.org/jira/browse/HADOOP-12916?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Xiaoyu Yao updated HADOOP-12916:
--------------------------------
Attachment: HADOOP-12916.00.patch
Attach a patch that
1) Separate the scheduler interface and implementation from FCQ
2) Allow DecayRpcScheduler to collect window based per priority level response
time metrics.
3) Allow optional backoff based on per priority level response time target
based backoff.
4) Additional unit test for response time target based backoff.
5) Additional unit test to ensure backward compatibility for current FCQ
deployment
With this refactor, the FCQ is purely a weighted priority call queue and the
DecayRpcScheduler is a fairness scheduler based on the callers' call count. The
weight priority call queue can be plugged with other priority based scheduler
for Qos.
> Allow different Hadoop IPC Call Queue throttling policies with FCQ/BackOff
> --------------------------------------------------------------------------
>
> Key: HADOOP-12916
> URL: https://issues.apache.org/jira/browse/HADOOP-12916
> Project: Hadoop Common
> Issue Type: Improvement
> Components: ipc
> Reporter: Xiaoyu Yao
> Assignee: Xiaoyu Yao
> Attachments: HADOOP-12916.00.patch
>
>
> Currently back off policy from HADOOP-10597 is hard coded to base on whether
> call queue is full. This ticket is open to allow flexible back off policies
> such as moving average of response time in RPC calls of different priorities.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
