Files go red and contain no data (see below), also still having client lock ups when a server is down. I don't see any errors in server logs (debug is on), but things just lock when server is down. When server is up, but glusterd is down clients work correctly. 007-08-01 16:05:04 E [afr.c:571:afr_getxattr_cbk] block-a-ds-afr: (path=/x9221-1.0 child=brick-a-ds) op_ret=-1 op_errno=61 2007-08-01 16:05:04 E [unify.c:335:unify_lookup] share-unify: /vs0_video0_16_2007-08-01.rs: Argument not right 2007-08-01 16:05:04 E [afr.c:571:afr_getxattr_cbk] block-a-ds-afr: (path=/lost+found child=brick-a-ds) op_ret=-1 op_errno=61 2007-08-01 16:05:04 E [afr.c:571:afr_getxattr_cbk] block-a-ds-afr: (path=/lost+found child=brick-a-ds) op_ret=-1 op_errno=61 ls -l total 41176 -rwxr-xr-x 1 root root 19625118 Aug 1 10:15 avframegrabber drwx------ 2 root root 4096 Aug 1 10:15 lost+found -rwxr-xr-x 1 root root 312 Aug 1 10:15 run -rwxrwxrwx 1 root root 225 Aug 1 10:52 setup ?--------- ? ? ? ? ? vs0_video0_16_2007-08-01.rs -rw-r--r-- 1 root root 7568990 Aug 1 16:04 vs0_video1_16_2007-08-01.rs -rw-r--r-- 1 root root 7279062 Aug 1 16:04 vs0_video2_16_2007-08-01.rs -rw-r--r-- 1 root root 7579511 Aug 1 16:04 vs0_video3_16_2007-08-01.rs drwxr-xr-x 2 root root 4096 Aug 1 14:52 x9221-1.0 [root@vs1 share]# 2007-08-01 16:05:21 E [afr.c:571:afr_getxattr_cbk] block-a-ds-afr: (path=/x9221-1.0 child=brick-a-ds) op_ret=-1 op_errno=61 2007-08-01 16:05:21 E [afr.c:571:afr_getxattr_cbk] block-a-ds-afr: (path=/x9221-1.0 child=brick-a-ds) op_ret=-1 op_errno=61 2007-08-01 16:05:21 E [unify.c:335:unify_lookup] share-unify: /vs0_video0_16_2007-08-01.rs: Argument not right 2007-08-01 16:05:21 E [afr.c:571:afr_getxattr_cbk] block-a-ds-afr: (path=/lost+found child=brick-a-ds) op_ret=-1 op_errno=61 2007-08-01 16:05:21 E [afr.c:571:afr_getxattr_cbk] block-a-ds-afr: (path=/lost+found child=brick-a-ds) op_ret=-1 op_errno=61 ><> Nathan Stratton CTO, Voila IP Communications nathan at robotics.net nathan at voilaip.com http://www.robotics.net http://www.voilaip.com On Wed, 1 Aug 2007, Harris Landgarten wrote: > Avati, > > What mount command are you referring to? > > Harris > > ----- Original Message ----- > From: "Anand Avati" <avati@xxxxxxxxxxxxx> > To: "Harris Landgarten" <harrisl@xxxxxxxxxxxxx> > Cc: "gluster-devel" <gluster-devel@xxxxxxxxxx> > Sent: Wednesday, August 1, 2007 1:03:58 PM (GMT-0500) America/New_York > Subject: Re: writebehind slowness > > Harris, > also please try with "--direct-io-mode=write-only" command line argument to the mount command. > > thanks, > avati > > > 2007/8/1 , Anand Avati < avati@xxxxxxxxxxxxx >: > > Harris, > the latest TLA has some cleanups about removing redundant synchronization locks, which can possible affect small files. can you please confirm if the performance hit on small files is seen with patch-402 as well? > > thanks! > > avati > > > 2007/8/1 , Harris Landgarten < harrisl@xxxxxxxxxxxxx >: > > time cp -av /usr/share/zoneinfo /mnt/glusterfs/test > > Harris > > ----- Original Message ----- > From: "Anand Avati" < avati@xxxxxxxxxxxxx > > To: "Harris Landgarten" < harrisl@xxxxxxxxxxxxx > > Cc: "gluster-devel" < gluster-devel@xxxxxxxxxx > > Sent: Wednesday, August 1, 2007 11:03:58 AM (GMT-0500) America/New_York > Subject: Re: writebehind slowness > > Harris, > with what application did you benchmark these number? > > avati > > > 2007/8/1 , Harris Landgarten < harrisl@xxxxxxxxxxxxx >: > > I am doing some very basic tests on 397. > > time cp -av /usr/share/zoneinfo /mnt/glusterfs/test > > with writebehind: 20s > > without writebehind: 12s > > Please check writebehind functionality. What test shows an improvement? > > > Harris > > > _______________________________________________ > Gluster-devel mailing list > Gluster-devel@xxxxxxxxxx > http://lists.nongnu.org/mailman/listinfo/gluster-devel > > > > -- > Anand V. Avati > > > > -- > Anand V. Avati > > > -- > Anand V. Avati > > > _______________________________________________ > Gluster-devel mailing list > Gluster-devel@xxxxxxxxxx > http://lists.nongnu.org/mailman/listinfo/gluster-devel >
