You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@synapse.apache.org by "Vanjikumaran Sivajothy (JIRA)" <ji...@apache.org> on 2015/07/22 20:52:04 UTC
[jira] [Updated] (SYNAPSE-998) Memory leak in VFS
[ https://issues.apache.org/jira/browse/SYNAPSE-998?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vanjikumaran Sivajothy updated SYNAPSE-998:
-------------------------------------------
Description:
Hi,
I have tested the VFS functionality with 500 proxies and each proxy pointed to different 500 folders. Then send around 300 concurrent HTTP request to each proxy. With the time i was able to notice memory growth and Synapse get Memory leak.
Therefore, i run the 300 concurrent HTTP request again the Synapse and killed client in the middle. I was able to clearly observe in Jconsole, memory was not released even when i try to perform GC manually.
To reproduce this, I have attached the python scripts that i have used to create Directories, Proxies and Requests.
was:
Hi,
I have tested the VFS functionality with 500 proxies and each proxy pointed to different 500 folders. Then send around 300 concurrent HTTP request to each proxy. With the time i was able to notice memory growth and Synapse get Memory leak.
Therefore, i run the 300 concurrent HTTP request again the Synapse and killed client in the middle. I was able to clearly observe in Jconsole, memory was not released even when i try to perform GC manually.
> Memory leak in VFS
> ------------------
>
> Key: SYNAPSE-998
> URL: https://issues.apache.org/jira/browse/SYNAPSE-998
> Project: Synapse
> Issue Type: Bug
> Reporter: Vanjikumaran Sivajothy
> Assignee: Hiranya Jayathilaka
> Priority: Critical
> Attachments: CreateDir.py, CreateProxy.py, HitHard.py
>
>
> Hi,
> I have tested the VFS functionality with 500 proxies and each proxy pointed to different 500 folders. Then send around 300 concurrent HTTP request to each proxy. With the time i was able to notice memory growth and Synapse get Memory leak.
> Therefore, i run the 300 concurrent HTTP request again the Synapse and killed client in the middle. I was able to clearly observe in Jconsole, memory was not released even when i try to perform GC manually.
> To reproduce this, I have attached the python scripts that i have used to create Directories, Proxies and Requests.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@synapse.apache.org
For additional commands, e-mail: dev-help@synapse.apache.org