cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jonathan Ellis (JIRA)" <j...@apache.org>
Subject [jira] Commented: (CASSANDRA-1869) Read and Write latencies should include stage queue time
Date Thu, 16 Dec 2010 22:17:04 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-1869?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12972245#action_12972245
] 

Jonathan Ellis commented on CASSANDRA-1869:
-------------------------------------------

You want the stage latency separate from waiting-in-queue latency (which can be easily derived
from queue size * latency).

If we need actually accurate, or at least histogram-accurate queue wait durations, we could
track that separately, but just adding it into the existing stage latency would be a step
backwards.

> Read and Write latencies should include stage queue time
> --------------------------------------------------------
>
>                 Key: CASSANDRA-1869
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-1869
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Stu Hood
>
> Our currently measured read and write latencies (as exposed on the CFS MBean) should
include the time spent queued in the read/write stages: only the active time is currently
measured, and this is much less useful.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message