[Gluster-devel] Problems with ioc again

[Dan Parsons]

I am once more having a problem with the ioc translator. Specifically,  
I upgraded the RAM in my gluster client from 4GB to 12GB. Under the  
4GB configuration, I had cache-size set to 2048MB. That was working  
alright. Once I upgraded the server to 12GB, I changed cache-size to  
4GB. After doing that, glusterfs proceeded to use every last drop of  
RAM, and all swap, on the system. Not right away of course, only after  
I read about, well, 12GB of data.

This behavior was repeated across all 33 of my gluster clients. Even  
back when it was set to 2GB, there were times when it would ignore  
that setting, and if you look through the mailing list, you'll find at  
least 2 people that had the same problem, and didn't get much support.

The ioc translator is an extremely essential part to the operation of  
my department. I really need this fixed ASAP- if there is anyone I can  
pay to do this, please contact me off-list, as it needs to be done ASAP.

[root at admin01 conf]# glusterfs --version
glusterfs 1.3.11 built on Aug 21 2008 11:26:38
Repository revision: glusterfs--mainline--2.5--patch-795

volume ioc
   type performance/io-cache
   subvolumes stripe0         # In this example it is 'client' you may  
have to change it according to your spec file.
   option page-size 1MB      # 128KB is default
   option cache-size 2048MB    # 32MB is default
   option force-revalidate-timeout 5 # 1second is default
   option priority *.psiblast:3,*.seq:2,*:1
end-volume



Dan Parsons