accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jared Winick (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ACCUMULO-1502) Add Trace Table docs to user manual
Date Wed, 21 Aug 2013 21:57:51 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-1502?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13746890#comment-13746890
] 

Jared Winick commented on ACCUMULO-1502:
----------------------------------------

I would like to help out here. I have starting playing around with tracing and have seen the
existing docs in docs/distributedTracing.html. I just have a few initial questions, technical
and administrative. 

1.) Should the documentation go in 1.5.1-SNAPSHOT? Seems like it would be reasonable to get
into the 1.5 code
2.) Is the Administration chapter of the user manual the appropriate place? Maybe right after
Monitoring?
3.) What server-side operations are traced by default? I have seen minor/major compactions
and GC.
4.) Sampling rate is currently left explicitly to the user via CountSampler, correct?
5.) Is there a good way to trace a write without flushing the BatchWriter? I know the actual
writes to the tablet servers are asynchronous from adding mutations so maybe it isn't possible.


Thanks for any feedback that can be provided.
                
> Add Trace Table docs to user manual
> -----------------------------------
>
>                 Key: ACCUMULO-1502
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-1502
>             Project: Accumulo
>          Issue Type: Sub-task
>          Components: docs, trace
>            Reporter: Sean Busbey
>             Fix For: 1.6.0
>
>
> Our feature list includes:
> {quote}
> TRACING
> It can be difficult to determine why some operations are taking longer than expected.
For example, you may be looking up items with very low latency, but sometimes the lookups
take much longer. Determining the cause of the delay is difficult because the system is distributed,
and the typical lookup is fast. Accumulo has been instrumented to record the time that various
operations take when tracing is turned on. The fact that tracing is enabled follows all the
requests made on behalf of the user throughout the distributed infrastructure of Accumulo,
and across all threads of execution.
> {quote}
> and the user manual makes no mention of the facility, its defaults, how to configure
it, nor how to leverage it in ops.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message