Tsukasa,
I did the same using the "GlusterFS High Availability Storage with
GlusterFS" document, and the performance was really bad.
After moving everything back to the client configs, the performance was
back.
My server config:
volume brick
type storage/posix # POSIX FS translator
option directory /home/export/webroot # Export this directory
end-volume
volume brick-ns
type storage/posix
option directory /home/export/namespace
end-volume
### Add network serving capability to above brick.
volume server
type protocol/server
option transport-type tcp/server # For TCP/IP transport
subvolumes brick brick-ns
option auth.ip.brick.allow 192.168.1.* # Allow access to "brick" volume
option auth.ip.brick-ns.allow 192.168.1.* # Allow access to "brick" volume
end-volume
My Client config:
### File: /etc/glusterfs-client.vol - GlusterFS Client Volume Specification
### Add client feature and attach to remote subvolume of server1
volume pnp-www1
type protocol/client
option transport-type tcp/client # for TCP/IP transport
option remote-host 192.168.1.159 # IP address of the remote brick
# option remote-port 6996 # default server port is 6996
option remote-subvolume brick # name of the remote volume
end-volume
### Add client feature and attach to remote subvolume of server2
volume pnp-www2
type protocol/client
option transport-type tcp/client # for TCP/IP transport
option remote-host 192.168.1.169 # IP address of the remote brick
# option remote-port 6996 # default server port is 6996
option remote-subvolume brick # name of the remote volume
end-volume
volume client-ns
type protocol/client
option transport-type tcp/client
option remote-host 192.168.1.159
option remote-subvolume brick-ns
end-volume
volume afr
type cluster/afr
subvolumes pnp-www1 pnp-www2
option replicate *:2
end-volume
### Add unify feature to cluster "server1" and "server2". Associate an
### appropriate scheduler that matches your I/O demand.
volume unify0
type cluster/unify
subvolumes afr
### ** Round Robin (RR) Scheduler **
option scheduler rr
### ** define a namespace server, which is not listed in 'subvolumes'
option namespace client-ns
end-volume
volume iot
type performance/io-threads
subvolumes unify0
option thread-count 8
end-volume
volume wb
type performance/write-behind
subvolumes iot
end-volume
volume ra
type performance/read-ahead
subvolumes wb
end-volume
volume ioc
type performance/io-cache
subvolumes ra
end-volume
Hope this helps you.
Guido
Tsukasa Morii wrote:
Hello.
I made an environment with GlusterFS by reference to the site below. But
disk peformance is not enough to be used for Web proxy storage. Could
somebody tell me the ways to make the peformance much better?
"GlusterFS High Availability Storage with GlusterFS"
http://www.gluster.org/docs/index.php/GlusterFS_High_Availability_Storage_with_GlusterFS
My goal and environment are as indicated below. Please let me know if
other informaton is needed.
[GOAL]
- Three web proxy servers keep a lot of web contens(html, css, js,
images) under gluster file system.
- Three servers can be used to build GlusterFS for storage.
- Automatic failover is necessary.
[ENVIRONMENT]
- 3 web proxy servers (CPU:2.4Ghz RAM:1GB)
- 3 gluster servers (CPU:2.4Ghz RAM:4GB)
- Most contents are less than 50k bytes.
_______________________________________________
Gluster-devel mailing list
Gluster-devel@xxxxxxxxxx
http://lists.nongnu.org/mailman/listinfo/gluster-devel
--
Met vriendelijke groet,
Guido Smit
ComLog B.V.
Televisieweg 133
1322 BE Almere
T. 036 5470500
F. 036 5470481
