Hi !
I have this Xen over Gluster deployment
3 Xen servers, 2 Gluster servers (2 bricks each one) replicated, 1 Gbit
dedicated network between all of them, about 200 Gbytes for 22 Xen
virtual machines.
All was fine until this action:
- Shutdown 2nd Gluster Server
- Move it (to another rack)
- Restart the server
At this point the auto-healing option activates and all 22 virtual
machines get blocked !
The process iowait had 100% of processor on 2dn gluster server.
At this moment I left just the 1st gluster server (2nd is disconnected
from network).
All servers have debian lenny with gluster 3.05.
Questions:
1- Is there a way to run de healing in background ?
2- Is there something wrong in my configs ? (attached at bottom)
Thanx in advance !
=============== Server config ====================
#
# Se define el primer brick
#
volume posix1
type storage/posix
option directory /disco1/01
option background-unlink yes # Se aconseja cuando el
filesystem contiene archivos de varios GB
end-volume
volume locks1
type features/posix-locks
option mandatory-locks on
subvolumes posix1
end-volume
volume brick1
type performance/io-threads
option thread-count 8 # Default es 16
subvolumes locks1
end-volume
#
# Se define el segundo brick
#
volume posix2
type storage/posix
option directory /disco1/02
option background-unlink yes # Se aconseja cuando el
filesystem contiene archivos de varios GB
end-volume
volume locks2
type features/posix-locks
option mandatory-locks on
subvolumes posix2
end-volume
volume brick2
type performance/io-threads
option thread-count 8 # Default es 16
subvolumes locks2
end-volume
#
# El servidor ofreciendo ambos bricks
#
volume server
type protocol/server
option transport-type tcp
option transport.socket.bind-address 10.253.2.8
option transport.socket.listen-port 7000
option auth.addr.brick1.allow *
option auth.addr.brick2.allow *
subvolumes brick1 brick2
end-volume
================ Client config ==================
#
# Se define primer brick en virgen
#
volume client1
type protocol/client
option transport-type tcp
option remote-host 10.253.2.9
option remote-port 7000
option remote-subvolume brick1
end-volume
#
# Se define segundo brick en virgen
#
volume client2
type protocol/client
option transport-type tcp
option remote-host 10.253.2.9
option remote-port 7000
option remote-subvolume brick2
end-volume
#
# Se define primer brick en mate
#
volume client3
type protocol/client
option transport-type tcp
option remote-host 10.253.2.8
option remote-port 7000
option remote-subvolume brick1
end-volume
#
# Se define segundo brick en mate
#
volume client4
type protocol/client
option transport-type tcp
option remote-host 10.253.2.8
option remote-port 7000
option remote-subvolume brick2
end-volume
#
# Se espejan los primeros brick de virgen y mate
#
volume server1
type cluster/replicate
subvolumes client1 client3
end-volume
#
# Se espejan los segundos brick de virgen y mate
#
volume server2
type cluster/replicate
subvolumes client2 client4
end-volume
#
# Se suman las replicaciones en un solo volumen completo
#
volume completo
type cluster/distribute
option min-free-disk 20%
option lookup-unhashed yes
subvolumes server1 server2
end-volume
#
# Se agregan opciones de performance
#
volume writebehind
type performance/write-behind
option cache-size 4MB
subvolumes completo
end-volume
volume iocache
type performance/io-cache
option cache-size 64MB
subvolumes writebehind
end-volume
--
Martin Eduardo Bradaschia
Intercomgi Argentina
_______________________________________________
Gluster-users mailing list
[email protected]
http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
