You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ambari.apache.org by "Oleksandr Diachenko (JIRA)" <ji...@apache.org> on 2013/10/22 16:31:45 UTC
[jira] [Updated] (AMBARI-3573) Configure Ganglia Web Server to use
KeepAlive connections
[ https://issues.apache.org/jira/browse/AMBARI-3573?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Oleksandr Diachenko updated AMBARI-3573:
----------------------------------------
Attachment: AMBARI-3573.patch
> Configure Ganglia Web Server to use KeepAlive connections
> ---------------------------------------------------------
>
> Key: AMBARI-3573
> URL: https://issues.apache.org/jira/browse/AMBARI-3573
> Project: Ambari
> Issue Type: Task
> Components: agent
> Affects Versions: 1.5.0
> Reporter: Oleksandr Diachenko
> Assignee: Oleksandr Diachenko
> Fix For: 1.5.0
>
> Attachments: AMBARI-3573.patch
>
>
> When performance profiling a 20 node cluster it was discovered that using KeepAlive http connections resulted in a 7% performance gain for a query that returned all host_component metrics.
> The GangliaPropertyProvider gets ganglia data by invoking the cgi script rrd.py. This results in a new HttpUrlConnection and input stream being obtained for each request to the ganglia back end.
> By default KeepAlive is enabled for Apache but we should ensure this is the case for the http server that provides rrd.py.
> For apache, this is configured in httpd.conf.
--
This message was sent by Atlassian JIRA
(v6.1#6144)