[
https://issues.apache.org/jira/browse/JENA-427?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13662277#comment-13662277
]
Andy Seaborne commented on JENA-427:
------------------------------------
I've started work on this JIRA.
1/ Added elapsed time per request (see above)
and next
2/ restructure the request lifecycle to make it possible to collect stats at
well defined points.
Introduce a per-request object (repurpose the current HttpAction but extend
scope from right at the start of the
HTTP request, not after deciding the operation type).
3/ Per dataset counters for every operation and whether successful or not.
4/ Per server counters
4/ Keep a finite buffer of recent request records.
Issues: consistency and parallelism. Counters will be AtomicLongs. Even so, a
separate total is needed because assing up the per service counters is not
necessarily going to give a consistent.
> Statistics for Fuseki usage.
> ----------------------------
>
> Key: JENA-427
> URL: https://issues.apache.org/jira/browse/JENA-427
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ, Fuseki, TDB
> Reporter: Andy Seaborne
> Assignee: Andy Seaborne
> Labels: mentor
>
> Between ARQ and Fuseki, there could be statistics on usage gathered for
> exposure via JMX.
> ARQ collects only a basic count.
> Fuseki could collect per dataset.
> What statistics would be useful? How would they be JMX-named?
> Statistics are useful for debugging and for long term system monitorings.
> Please add suggestions to the JIRA.
--
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
