[Gluster-users] GFS 2.0.2 Stabilitiy issues

[Shehjar Tikoo]

Scandic wrote:
> I'm getting stability issues with 2.0x series. The log entries are:
> 
> [2009-06-07 21:31:09] E [client-protocol.c:292:call_bail] Node2: bailing 
> out frame LOOKUP(32) frame sent = 2009-06-07 21:01:02. frame-timeout = 1800
> 
> Currently this a test cluster so there's not much traffic going on. 
> However, the GlusterFS mount isn't stable for more than 48 hours. Any 
> ideas?
> 
> Then the GlusterFS mount diappears.
> 
> My configuration:
> 
> volume posix
>  type storage/posix
>  option directory /home/export
> end-volume
> 
> volume locks
>  type features/locks
>  subvolumes posix
> end-volume
> 
> volume brick
>  type performance/io-threads
>  option autoscaling yes
>  subvolumes locks
> end-volume
> 

First, try without autoscaling.
Then, if you need autoscaling, use a non-default thread count by using:

option max-threads 32

Currently, the default is 256. This has already been reduced by a commit
in the repository.

-Shehjar

> volume server
>  type protocol/server
>  option transport-type ib-verbs
>  option transport.ib-verbs.listen-port 6990
>  option auth.addr.brick.allow *
>  subvolumes brick
> end-volume
> 
> volume Node1
>  type protocol/client
>  option transport-type ib-verbs
>  option remote-port 6990
>  option remote-host Node1
>  option remote-subvolume brick
> end-volume
> 
> volume Node2
>  type protocol/client
>  option transport-type ib-verbs
>  option remote-port 6990
>  option remote-host Node2
>  option remote-subvolume brick
> end-volume
> 
> volume nufa
>  type cluster/replicate
>  option scheduler nufa
>  option read-subvolume brick
>  subvolumes Node1 Node2
> end-volume
> 
> volume writebehind
>   type performance/write-behind
>   option cache-size 8MB
>   subvolumes nufa
> end-volume
> 
> volume cache
>   type performance/io-cache
>   option cache-size 8MB
>   subvolumes writebehind
> end-volume
> 
> 
> 
> 
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users
>