cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Rich Rein (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-8381) CFStats should record keys of largest N requests for time interval
Date Wed, 26 Nov 2014 21:24:12 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-8381?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14226822#comment-14226822
] 

Rich Rein commented on CASSANDRA-8381:
--------------------------------------

It would be extremely useful to have 1 or a few recent key values for each step of the histogram.
This would allow developers to see partitioning sizes as a side affect of key values and frequency
correlated with partition sizes.

But the worse case sizes mentioned by matt is critical.

> CFStats should record keys of largest N requests for time interval
> ------------------------------------------------------------------
>
>                 Key: CASSANDRA-8381
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-8381
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Matt Stump
>            Priority: Critical
>
> Isolating the problem partition for a CF is right now incredibly difficult. If we could
keep the primary key of the largest N read or write requests for the pervious interval or
since counter has been cleared it would be extremely useful.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message