I've just setup a simple gluster storage system on Centos 5.2 x64 w/ gluster
1.3.10
I have three storage bricks and one client
Everytime i run iozone across this setup, i seem to get a bad file
descriptor around the 4k mark.
Any thoughts why? I'm sure more info is wanted, i'm just not sure what else
to include at this point.
thanks
[root at green gluster]# cat /opt/gluster/etc/glusterfs/gluster.client
volume remote1
type protocol/client
option transport-type tcp/client
option remote-host n1
option remote-subvolume brick
end-volume
volume remote2
type protocol/client
option transport-type tcp/client
option remote-host n2
option remote-subvolume brick
end-volume
volume remote3
type protocol/client
option transport-type tcp/client
option remote-host n3
option remote-subvolume brick
end-volume
volume stripe0
type cluster/stripe
option block-size *:1MB
subvolumes remote1 remote2 remote3
end-volume
[root at green gluster]# cat /opt/gluster/etc/glusterfs/gluster.server
volume brick
type storage/posix
option directory /gluster
end-volume
volume server
type protocol/server
option transport-type tcp/server
option auth.ip.brick.allow *
subvolumes brick
end-volume
--- iozone snipped...
4096 4 4109 4228 7336 10409 4245 4209
4163 4232 4178 4098 4215 10218 6782
4096 8 5951 6093 11445 11450 6096 6085
6028 6104 6054 5935 6077 11409 6792
4096 16 7609 7757 11395 7353 7825 7747
7705 7793 7776 7595 7753 11396 7078
4096 32 8915 9074 11313 10268 9166 9074
8977 9118 9127 8922 9083
Error freading block 62 9d800000
read: File descriptor in bad state
[root at green gluster]#
[root at green gluster]#
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://zresearch.com/pipermail/gluster-users/attachments/20080801/c710d79c/attachment.htm
