You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@sling.apache.org by "Bertrand Delacretaz (JIRA)" <ji...@apache.org> on 2016/08/31 08:50:20 UTC
[jira] [Comment Edited] (SLING-5410) Identify component where
important Metrics can be collected
[ https://issues.apache.org/jira/browse/SLING-5410?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15087187#comment-15087187 ]
Bertrand Delacretaz edited comment on SLING-5410 at 8/31/16 8:50 AM:
---------------------------------------------------------------------
Monitoring does add overhead in the 1E-7 or 1E-8s range so don't try and instrument things that would be better done with a profiler. Instrument areas of the code that give a high level view, and be prepared to iterate as more is learnt from production or developer testing.
Some potential first targets are [1], to monitor the performance of the core Sling Engine in respect of request processing, component processing and request dispatching. This will not expose output stream processing/rewriting but should give some initial information.
handleSecurity in [2], should give some insight into the general performance of the authN stack.
I suspect there are similar locations in Sling Event and Discovery that may be of value to monitor.
1 https://github.com/apache/sling/blob/trunk/bundles/engine/src/main/java/org/apache/sling/engine/impl/SlingRequestProcessorImpl.java
2 https://github.com/apache/sling/blob/7c4a53755aed1211c9af313a3973cd2543a7bbe0/bundles/engine/src/main/java/org/apache/sling/engine/impl/SlingHttpContext.java#L111
was (Author: ianeboston):
Monitoring does add overhead in the 1E-7 or 1E-8s range do don't try and instrument things that would be better done with a profiler. Instrument areas of the code that give a high level view, and be prepared to iterate as more is learnt from production or developer testing.
Some potential first targets are [1], to monitor the performance of the core Sling Engine in respect of request processing, component processing and request dispatching. This will not expose output stream processing/rewriting but should give some initial information.
handleSecurity in [2], should give some insight into the general performance of the authN stack.
I suspect there are similar locations in Sling Event and Discovery that may be of value to monitor.
1 https://github.com/apache/sling/blob/trunk/bundles/engine/src/main/java/org/apache/sling/engine/impl/SlingRequestProcessorImpl.java
2 https://github.com/apache/sling/blob/7c4a53755aed1211c9af313a3973cd2543a7bbe0/bundles/engine/src/main/java/org/apache/sling/engine/impl/SlingHttpContext.java#L111
> Identify component where important Metrics can be collected
> -----------------------------------------------------------
>
> Key: SLING-5410
> URL: https://issues.apache.org/jira/browse/SLING-5410
> Project: Sling
> Issue Type: Task
> Components: Engine
> Reporter: Chetan Mehrotra
> Assignee: Chetan Mehrotra
>
> With SLING-4080 we now have support for collecting various kinds of metrics in performant way.
> As a next step we would need to identify those parts of Sling engine and request processing logic where we can collect important metrics. Such metric should help
> # Sling system administrator to monitor key aspects of Sling
> # Allow us as developers to get insight into critical parts and help in doing further optimization
> Once identified we can then add required metric collection logic to those parts
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)