Markus Fröhlich
2011-Jan-27 16:03 UTC
[Gluster-users] sometimes connection errors - glusterfs limit?
hi!
problem description:
1-3 times a day glusterfs seems to hang - like freezed - network traffic is a
few secounds again
zero, because there is no glusterfs communication.
in the log of the servers, that clients got disconnected and a little bit later
connected again
but the network traffic seems okay and is not at limit - there are no errors on
the networkinterfaces.
we also checked cables and switchports on the switches.
we think that glusterfs in combination with the I/O is the bottleneck here.
average we can say the storage setup is in use with 2/3 read and 1/3 write
operations.
is there a limitation of volumes, clients, mounts etc. in the glusterfs code?
has some one made similar experience or troubles with such a setup?
is it possible that to much clients are connected to few servers?
has any one some tip for us?
-------------------------------------
our setup:
two glusterfs servers:
* areca raid controller with raid5 setup
* 3 LUNs each 11 TB with 70%-99% in use with ext3 formated
* SLES11 x86_64
* glusterfs V 3.0.7
45 - 50 glusterfs client servers:
* SLES10, SLES11, SLES11 SP1
* glusterfs V 3.0.7
* all VOLs client replicated to the both glusterfs servers
-----------------------------------
volume files:
server export vols look like this - some options differ:
--
volume posix
type storage/posix
option directory /gluster-storage/projekte/ksc/
option background-unlink yes
end-volume
volume locks
type features/locks
subvolumes posix
end-volume
volume ksc
type performance/io-threads
option thread-count 16
subvolumes locks
end-volume
volume server
type protocol/server
option transport-type tcp
option transport.socket.listen-port 7025
option auth.addr.ksc.allow 10.0.1.*
subvolumes ksc
end-volume
--
volume posix
type storage/posix
option directory /gluster-storage/projekte/hosting/
option o-direct enable
option background-unlink yes
end-volume
volume locks
type features/locks
subvolumes posix
end-volume
volume hosting2
type performance/io-threads
option thread-count 16
subvolumes locks
end-volume
volume server
type protocol/server
option transport-type tcp
option transport.socket.listen-port 7005
option auth.addr.hosting2.allow 10.0.1.*
subvolumes hosting2
end-volume
--
client repl. mount VOL files:
volume vgfs-01-001-ksc
type protocol/client
option transport-type tcp
option remote-host vgfs-01-001
option remote-port 7025
option ping-timeout 10
option remote-subvolume ksc
end-volume
# distribute
volume distribute1-ksc
type cluster/distribute
option lookup-unhashed auto
option min-free-disk 5%
subvolumes vgfs-01-001-ksc
end-volume
volume vgfs-01-002-ksc
type protocol/client
option transport-type tcp
option remote-host vgfs-01-002
option remote-port 7025
option ping-timeout 10
option remote-subvolume ksc
end-volume
# distribute
volume distribute2-ksc
type cluster/distribute
option lookup-unhashed auto
option min-free-disk 5%
subvolumes vgfs-01-002-ksc
end-volume
volume ksc-data-replicate
type cluster/replicate
subvolumes distribute1-ksc distribute2-ksc
end-volume
volume iocache
type performance/io-cache
option cache-size 64MB #1GB supported
option cache-timeout 1
subvolumes ksc-data-replicate
end-volume
volume quick-read
type performance/quick-read
# option cache-timeout 10 (1 second)
# option max-file-size 1048576 (64Kb)
subvolumes iocache
end-volume
volume trace
type debug/trace
subvolumes quick-read
# option include open,close,create,readdir,opendir,closedir
# option exclude lookup,read,write
end-volume
--
volume vgfs-01-001-hosting
type protocol/client
option transport-type tcp
option remote-host vgfs-01-001
option remote-port 7005
option ping-timeout 20
option remote-subvolume hosting
end-volume
volume vgfs-01-002-hosting
type protocol/client
option transport-type tcp
option remote-host vgfs-01-002
option remote-port 7005
option ping-timeout 20
option remote-subvolume hosting
end-volume
# distribute
volume distribute1-hosting
type cluster/distribute
option lookup-unhashed yes
option min-free-disk 5%
subvolumes vgfs-01-001-hosting
end-volume
# distribute
volume distribute2-hosting
type cluster/distribute
option lookup-unhashed yes
option min-free-disk 5%
subvolumes vgfs-01-002-hosting
end-volume
volume backup-data-replicate
type cluster/replicate
subvolumes distribute1-hosting distribute2-hosting
subvolumes distribute2-hosting
end-volume
volume readahead
type performance/read-ahead
option page-count 16 # cache per file = (page-count x
page-size)
subvolumes backup-data-replicate
end-volume
volume iocache
type performance/io-cache
option cache-size 1024MB #1GB supported
option cache-timeout 1
subvolumes readahead
end-volume
volume iothreads
type performance/io-threads
option thread-count 6 # default is 16
subvolumes iocache
end-volume
volume quickread
type performance/quick-read
option cache-timeout 30
option max-file-size 1024000
subvolumes iothreads
end-volume
--
regards
markus
--
Mit freundlichen Gr?ssen
Markus Fr?hlich
Techniker
________________________________________________________
Xidras GmbH
Stockern 47
3744 Stockern
Austria
Tel: +43 (0) 2983 201 30503
Fax: +43 (0) 2983 201 305039
Email: markus.froehlich at xidras.com
Web: http://www.xidras.com
FN 317036 f | Landesgericht Krems | ATU64485024
________________________________________________________________________________
VERTRAULICHE INFORMATIONEN!
Diese eMail enth?lt vertrauliche Informationen und ist nur f?r den
berechtigten Empf?nger bestimmt. Wenn diese eMail nicht f?r Sie bestimmt
ist, bitten wir Sie, diese eMail an uns zur?ckzusenden und anschlie?end
auf Ihrem Computer und Mail-Server zu l?schen. Solche eMails und Anlagen
d?rfen Sie weder nutzen, noch verarbeiten oder Dritten zug?nglich
machen, gleich in welcher Form.
Wir danken f?r Ihre Kooperation!
CONFIDENTIAL!
This email contains confidential information and is intended for the
authorised recipient only. If you are not an authorised recipient,
please return the email to us and then delete it from your computer
and mail-server. You may neither use nor edit any such emails including
attachments, nor make them accessible to third parties in any manner
whatsoever.
Thank you for your cooperation
________________________________________________________________________________
Burnash, James
2011-Jan-27 16:57 UTC
[Gluster-users] sometimes connection errors - GlusterFS limit?
Hello.
I have experienced this situation with the 3.0.4 release of Glusterfs - it was
related to a bug that had to do with recursive file deletions (in my case).
That bug has been fixed in 3.1.1 which is what I am currently running.
Can you give us your Glusterfs version, and a copy of your volume files for
server and client?
That would help us to help you.
Thanks,
James
-----Original Message-----
From: gluster-users-bounces at gluster.org [mailto:gluster-users-bounces at
gluster.org] On Behalf Of Markus Fr?hlich
Sent: Thursday, January 27, 2011 11:03 AM
To: gluster-users at gluster.org
Subject: [Gluster-users] sometimes connection errors - glusterfs limit?
hi!
problem description:
1-3 times a day glusterfs seems to hang - like freezed - network traffic is a
few secounds again
zero, because there is no glusterfs communication.
in the log of the servers, that clients got disconnected and a little bit later
connected again
but the network traffic seems okay and is not at limit - there are no errors on
the networkinterfaces.
we also checked cables and switchports on the switches.
we think that glusterfs in combination with the I/O is the bottleneck here.
average we can say the storage setup is in use with 2/3 read and 1/3 write
operations.
is there a limitation of volumes, clients, mounts etc. in the glusterfs code?
has some one made similar experience or troubles with such a setup?
is it possible that to much clients are connected to few servers?
has any one some tip for us?
-------------------------------------
our setup:
two glusterfs servers:
* areca raid controller with raid5 setup
* 3 LUNs each 11 TB with 70%-99% in use with ext3 formated
* SLES11 x86_64
* glusterfs V 3.0.7
45 - 50 glusterfs client servers:
* SLES10, SLES11, SLES11 SP1
* glusterfs V 3.0.7
* all VOLs client replicated to the both glusterfs servers
-----------------------------------
volume files:
server export vols look like this - some options differ:
--
volume posix
type storage/posix
option directory /gluster-storage/projekte/ksc/
option background-unlink yes
end-volume
volume locks
type features/locks
subvolumes posix
end-volume
volume ksc
type performance/io-threads
option thread-count 16
subvolumes locks
end-volume
volume server
type protocol/server
option transport-type tcp
option transport.socket.listen-port 7025
option auth.addr.ksc.allow 10.0.1.*
subvolumes ksc
end-volume
--
volume posix
type storage/posix
option directory /gluster-storage/projekte/hosting/
option o-direct enable
option background-unlink yes
end-volume
volume locks
type features/locks
subvolumes posix
end-volume
volume hosting2
type performance/io-threads
option thread-count 16
subvolumes locks
end-volume
volume server
type protocol/server
option transport-type tcp
option transport.socket.listen-port 7005
option auth.addr.hosting2.allow 10.0.1.*
subvolumes hosting2
end-volume
--
client repl. mount VOL files:
volume vgfs-01-001-ksc
type protocol/client
option transport-type tcp
option remote-host vgfs-01-001
option remote-port 7025
option ping-timeout 10
option remote-subvolume ksc
end-volume
# distribute
volume distribute1-ksc
type cluster/distribute
option lookup-unhashed auto
option min-free-disk 5%
subvolumes vgfs-01-001-ksc
end-volume
volume vgfs-01-002-ksc
type protocol/client
option transport-type tcp
option remote-host vgfs-01-002
option remote-port 7025
option ping-timeout 10
option remote-subvolume ksc
end-volume
# distribute
volume distribute2-ksc
type cluster/distribute
option lookup-unhashed auto
option min-free-disk 5%
subvolumes vgfs-01-002-ksc
end-volume
volume ksc-data-replicate
type cluster/replicate
subvolumes distribute1-ksc distribute2-ksc
end-volume
volume iocache
type performance/io-cache
option cache-size 64MB #1GB supported
option cache-timeout 1
subvolumes ksc-data-replicate
end-volume
volume quick-read
type performance/quick-read
# option cache-timeout 10 (1 second)
# option max-file-size 1048576 (64Kb)
subvolumes iocache
end-volume
volume trace
type debug/trace
subvolumes quick-read
# option include open,close,create,readdir,opendir,closedir
# option exclude lookup,read,write
end-volume
--
volume vgfs-01-001-hosting
type protocol/client
option transport-type tcp
option remote-host vgfs-01-001
option remote-port 7005
option ping-timeout 20
option remote-subvolume hosting
end-volume
volume vgfs-01-002-hosting
type protocol/client
option transport-type tcp
option remote-host vgfs-01-002
option remote-port 7005
option ping-timeout 20
option remote-subvolume hosting
end-volume
# distribute
volume distribute1-hosting
type cluster/distribute
option lookup-unhashed yes
option min-free-disk 5%
subvolumes vgfs-01-001-hosting
end-volume
# distribute
volume distribute2-hosting
type cluster/distribute
option lookup-unhashed yes
option min-free-disk 5%
subvolumes vgfs-01-002-hosting
end-volume
volume backup-data-replicate
type cluster/replicate
subvolumes distribute1-hosting distribute2-hosting
subvolumes distribute2-hosting
end-volume
volume readahead
type performance/read-ahead
option page-count 16 # cache per file = (page-count x
page-size)
subvolumes backup-data-replicate
end-volume
volume iocache
type performance/io-cache
option cache-size 1024MB #1GB supported
option cache-timeout 1
subvolumes readahead
end-volume
volume iothreads
type performance/io-threads
option thread-count 6 # default is 16
subvolumes iocache
end-volume
volume quickread
type performance/quick-read
option cache-timeout 30
option max-file-size 1024000
subvolumes iothreads
end-volume
--
regards
markus
--
Mit freundlichen Gr?ssen
Markus Fr?hlich
Techniker
________________________________________________________
Xidras GmbH
Stockern 47
3744 Stockern
Austria
Tel: +43 (0) 2983 201 30503
Fax: +43 (0) 2983 201 305039
Email: markus.froehlich at xidras.com
Web: http://www.xidras.com
FN 317036 f | Landesgericht Krems | ATU64485024
________________________________________________________________________________
VERTRAULICHE INFORMATIONEN!
Diese eMail enth?lt vertrauliche Informationen und ist nur f?r den
berechtigten Empf?nger bestimmt. Wenn diese eMail nicht f?r Sie bestimmt
ist, bitten wir Sie, diese eMail an uns zur?ckzusenden und anschlie?end
auf Ihrem Computer und Mail-Server zu l?schen. Solche eMails und Anlagen
d?rfen Sie weder nutzen, noch verarbeiten oder Dritten zug?nglich
machen, gleich in welcher Form.
Wir danken f?r Ihre Kooperation!
CONFIDENTIAL!
This email contains confidential information and is intended for the
authorised recipient only. If you are not an authorised recipient,
please return the email to us and then delete it from your computer
and mail-server. You may neither use nor edit any such emails including
attachments, nor make them accessible to third parties in any manner
whatsoever.
Thank you for your cooperation
________________________________________________________________________________
_______________________________________________
Gluster-users mailing list
Gluster-users at gluster.org
http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
DISCLAIMER:
This e-mail, and any attachments thereto, is intended only for use by the
addressee(s) named herein and may contain legally privileged and/or confidential
information. If you are not the intended recipient of this e-mail, you are
hereby notified that any dissemination, distribution or copying of this e-mail,
and any attachments thereto, is strictly prohibited. If you have received this
in error, please immediately notify me and permanently delete the original and
any copy of any e-mail and any printout thereof. E-mail transmission cannot be
guaranteed to be secure or error-free. The sender therefore does not accept
liability for any errors or omissions in the contents of this message which
arise as a result of e-mail transmission.
NOTICE REGARDING PRIVACY AND CONFIDENTIALITY Knight Capital Group may, at its
discretion, monitor and review the content of all e-mail communications.
http://www.knight.com