[Gluster-users] Quickread Translator Memory Leak?

Benjamin Long benjamin.long at longbros.com
Fri Mar 19 14:16:49 UTC 2010


On Friday 19 March 2010 03:17:29 am Vijay Bellur wrote:
> Benjamin Long wrote:
> >> Do you have details of GlusterFS's memory usage (Resident Memory and
> >> percentage of memory used) at the instant when the oom condition was
> >> observed?
> >
> > Yep. It's a VM with 1GB of ram. It runs nothing but gluster, rsync, and
> > ssh. I saw glusterfs using 97% of the ram just before it died. All the
> > swap was used up too.
> >
> > Here's the output of top about 10 min before that:
> >   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
> >  2868 root      20   0 55664 2356  648 S   14  0.2   0:23.28 rsync
> >  2239 root      20   0  933m 752m 1300 R    5 74.9   0:12.06 glusterfs
> 
> Can you please provide us the GlusterFS version that you are running and
> the client volume file? If it is 3.0.0 and you are using configuration
> generated by glusterfs-volgen in a system that does not have too much
> RAM, this is a possibility.
> 
> Regards,
> Vijay
> 

I'm currently running 3.0.2. Here is the client volfile for the user home 
directories. It was originaly created by volgen, but I made some changes.

# RAID 1
# TRANSPORT-TYPE tcp
volume 10.10.2.29-1
    type protocol/client
    option transport-type tcp
    option remote-host 10.10.2.29
    option transport.socket.nodelay on
    option transport.remote-port 6997
    option remote-subvolume home-ds-locks-io-filter
end-volume

volume 10.10.2.30-1
    type protocol/client
    option transport-type tcp
    option remote-host 10.10.2.30
    option transport.socket.nodelay on
    option transport.remote-port 6997
    option remote-subvolume home-ds-locks-io-filter
end-volume

volume mirror-0
    type cluster/replicate
    subvolumes 10.10.2.29-1 10.10.2.30-1
    option metadata-change-log on
#    option data-lock-server-count 2
#    option metadata-lock-server-count 2
#    option entry-lock-server-count 2
end-volume

volume writebehind
    type performance/write-behind
    option cache-size 4MB
    subvolumes mirror-0
end-volume

volume readahead
    type performance/read-ahead
    option page-count 4
    subvolumes writebehind
end-volume

volume iocache
    type performance/io-cache
    option cache-size 100MB
    option cache-timeout 1
    subvolumes readahead
end-volume

volume quickread
    type performance/quick-read
    option cache-timeout 1
    option max-file-size 64kB
    subvolumes iocache
end-volume

volume statprefetch
    type performance/stat-prefetch
    subvolumes quickread
end-volume
-- 
Benjamin Long



More information about the Gluster-users mailing list