Re: [Gluster-devel] rdiff-backup to glusterfs share doesn't work at all

Hello, I'm sorry for the noise. The problem seems to be completely unrelated to GlusterFS. An upgrade of lib pyxattr 0.2 to 0.4 fixed everything. So the problem was between rdiff-backup 1.2.2 and pyxattr 0.2. I'm sorry for blaming GlusterFS without proper debugging. Hannes Dor

Re: [Gluster-devel] rdiff-backup to glusterfs share doesn't work at all

Hello, it seems to be a similar problem to what I've had with rsync once: http://www.mail-archive.com/gluster-devel@nongnu.org/msg03361.html The file seems already renamed when the xattr call arrives. Hannes Dorbath wrote: > Hello, > > I'm using GlusterFS 1.3.12. I'

[Gluster-devel] rdiff-backup to glusterfs share doesn't work at all

not match source I tried to remove ra/wb/iothreads translators, but it didn't help. The share is a unify volume from 3 discs. Any idea how to solve that? Thanks in advance. -- Best regards, Hannes Dorbath ___ Gluster-devel mailing list Glus

Re: mod_glusterfs (was Re: [Gluster-devel] Hammering start() calls)

ve this problem at all, as it loads all files once on startup. -- Best regards, Hannes Dorbath ___ Gluster-devel mailing list Gluster-devel@nongnu.org http://lists.nongnu.org/mailman/listinfo/gluster-devel

Re: mod_glusterfs (was Re: [Gluster-devel] Hammering start() calls)

e option cache-size 64MB option page-size 128KB option force-revalidate-timeout 600 subvolumes wb end-volume -- Best regards, Hannes Dorbath ___ Gluster-devel mailing list Gluster-devel@nongnu.org http://lists.nongnu.org/mailman/listinfo/gluster-devel

Re: [Gluster-devel] Hammering start() calls

. Would booster help me? Is it fuse causing the latency? Thanks in advance. -- Best regards, Hannes Dorbath ___ Gluster-devel mailing list Gluster-devel@nongnu.org http://lists.nongnu.org/mailman/listinfo/gluster-devel

[Gluster-devel] Hammering start() calls

someone come up with a recommendation for this workload? No unify, no AFR, pure NFS replacement, tons of very small files. Thanks in advance. -- Best regards, Hannes Dorbath ___ Gluster-devel mailing list Gluster-devel@nongnu.org http

Re: [Gluster-devel] Re: High CPU usage -- any way to reduce cpu load?

base on a network file systems is in general considered a very bad idea. Look into block level based replication for a DBMS (DRBD, ENBD, iSCSI, AoE, FC, etc). -- Best regards, Hannes Dorbath ___ Gluster-devel mailing list Gluster-devel@nongnu.org

Re: [Gluster-devel] Unify losing files with rsync

Anand Avati wrote: Hannes, thanks for the data. We are working on it. I just like to confirm the fix is working fine. Thanks. -- Best regards, Hannes Dorbath ___ Gluster-devel mailing list Gluster-devel@nongnu.org http://lists.nongnu.org/mailman

Re: [Gluster-devel] Unify losing files with rsync

Hannes Dorbath wrote: Hannes Dorbath wrote: http://theendofthetunnel.de/rsync.txt I just noticed that the files that fail have a dot prefixed and an extension add.. like ".foo.mp3.EJWWp1" -- is the a temporary name rsync uses? Could that be a race condition because of io-threa

Re: [Gluster-devel] Unify losing files with rsync

Hannes Dorbath wrote: http://theendofthetunnel.de/rsync.txt I just noticed that the files that fail have a dot prefixed and an extension add.. like ".foo.mp3.EJWWp1" -- is the a temporary name rsync uses? Could that be a race condition because of io-threads? -- Best regar

[Gluster-devel] Unify losing files with rsync

hand to debug this I'd really appreciate it. -- Best regards, Hannes Dorbath ___ Gluster-devel mailing list Gluster-devel@nongnu.org http://lists.nongnu.org/mailman/listinfo/gluster-devel