perfSONAR User Q&A and Other Discussion

[Shawn McKee <>]

Hi Darryl,

This is a known issue and will be fixed once perfSONAR v3.4.2 is release (hopefully soon!)

We (WLCG/OSG) have a FAQ item on this at: https://twiki.opensciencegrid.org/bin/view/Documentation/TroubleFAQPS#My_perfSONAR_disks_are_filling_u

Let me know if you have any questions or problems.

Shawn

On Thu, Jan 29, 2015 at 12:41 PM, Darryl K Wohlt <> wrote:

My OWAMP host on the USCMS T2 mesh keeps running out of disk space in its root partition, which essentially kills the machine.  As far as I can determine, the main offender is the /var/log/perfsonar/regular_testing.log, which collects over 3GB (>155K messages/hour) of information each day, and of course the logs are retained for 7 days.  The disk space runs out within a few days, and I have to remove the older regular_testing.log.[1-7] files to get things going again.  I’ve even set up an MRTG job to monitor it and send me email when it hits 95%.

 

PSONAR4 Disk Usage

System:	PSONAR4
Maintainer:
Resource:	UCD-SNMP-MIB:dskTable

 

`Weekly' Graph (30 Minute Average)

Max root	100.0 %		Average root	79.1 %		Current root	62.0 %
Max boot	30.0 %		Average boot	30.0 %		Current boot	30.0 %

 

Obviously in this case I missed the email.  :-}

 

Has anyone else noticed something like this?  Is there a parameter I can set to reduce the number of retained logs or reduce the amount of information logged?

 

Thanks,

Darryl

 

Darryl K. Wohlt

Network Architect I

 

CCD/NCS/Network Services

Fermi National Accelerator Laboratory

P.O. Box 500, MS 368

Batavia, Illinois 60510

USA

 

630 840 2901 office

630 945 5687  mobile

www.fnal.gov