[Gluster-users] client side profiling

Computerisms Corporation bob at computerisms.ca
Fri Aug 21 04:44:35 UTC 2020

Hi List,

I am still struggling with my setup.  One server is working reasonably 
well for serving websites, but serving sites from the 2nd server is 
still using excessive amounts of cpu; a bit of which is gluster, but 
most of which is apache.

Gluster docs mentions client-side-profiling:



"In short, use client-side profiling for understanding "why is my 
application unresponsive"?"

Great, I think this is what I want.  instructions are:

*gluster volume profile your-volume start
*setfattr -n trusted.io-stats-dump -v /tmp/io-stats-pre.txt /your/mountpoint
*This will generate the specified file on the client


root at moogle:/usr/src/gluster-profile-analysis-master# gluster vol 
profile webisms start
Profile on Volume webisms is already started
root at moogle:/usr/src/gluster-profile-analysis-master# setfattr -n 
trusted.io-stats-dump -v /tmp/stats.txt /Computerisms
root at moogle:/usr/src/gluster-profile-analysis-master# ls /tmp/stats.txt
ls: cannot access '/tmp/stats.txt': No such file or directory

thought for sure I am doing something wrong, so I had a look at the 
gvp-client.sh script, and it appears I am doing the command correctly, 
there is just no output file.  Am I missing something?  or is this an 
outdated methodology that no longer works?

Bob Miller
Cell: 867-334-7117
Office: 867-633-3760
Office: 867-322-0362

More information about the Gluster-users mailing list