Can't start a volume.
[root at sfsccl03 ~]# gluster volume start brick1
brick: sfsccl03:/data/brick-sdc2/glusterfs/dht, path creation failed,
reason: No such file or directory
But ...
[root at sfsccl03 ~]# ls -alF /data/brick-sdc2/glusterfs
total 0
drwxr-xr-x 4 root root 27 Dec 2 13:00 ./
drwxr-xr-x 4 root root 107 Jul 5 11:55 ../
drwxrwxrwt 7 root root 61 Sep 15 11:35 dht/
drwxr-xr-x 2 root root 6 Dec 2 13:00 dht2/
So it is there.
[root at sfsccl03 ~]# ls -alF /data/brick-sdc2/glusterfs/dht
total 128
drwxrwxrwt 7 root root 61 Sep 15 11:35 ./
drwxr-xr-x 4 root root 27 Dec 2 13:00 ../
drwxr-xr-x 1230 root root 65536 Oct 24 09:14 equity/
drwxr-xr-x 1740 oracle root 65536 Nov 30 23:33 opra/
drwxr-xr-x 35 oracle oinstall 501 Jul 9 17:07 tag/
drwxr-xr-x 11 root root 126 Jul 1 08:51 taq/
drwxr-xr-x 2 root root 34 Jul 11 19:44 test/
and it is readable.
More info:
[root at sfsccl03 ~]# gluster volume info brick1
Volume Name: brick1
Type: Distribute
Status: Stopped
Number of Bricks: 4
Transport-type: tcp
Bricks:
Brick1: sfsccl01:/data/glusterfs/dht
Brick2: sfsccl02:/data/glusterfs/dht
Brick3: sfsccl03:/data/brick-sdc2/glusterfs/dht
Brick4: sfsccl03:/data/brick-sdd2/glusterfs/dht
[root at sfsccl03 ~]# gluster peer status
Number of Peers: 2
Hostname: sfsccl02
Uuid: 6e72d1a8-bdeb-4bfb-806c-7fa8b98cb697
State: Peer in Cluster (Connected)
Hostname: sfsccl01
Uuid: 116197cd-5dfe-4881-85ad-5de2be484ba6
State: Peer in Cluster (Connected)
a volume reset doesn't help.
[root at sfsccl03 ~]# gluster volume reset brick1
reset volume successful
[root at sfsccl03 ~]# gluster volume start brick1
brick: sfsccl03:/data/brick-sdc2/glusterfs/dht, path creation failed,
reason: No such file or directory
New volume creation also fails.
[root at sfsccl03 ~]# gluster volume create brick2 transport tcp
sfsccl01:/data/glusterfs/dht2 sfsccl03:/data/brick-sdc2/glusterfs/dht2
sfsccl02:/data/glusterfs/dht2 sfsccl03:/data/brick-sdd2/glusterfs/dht2
brick: sfsccl03:/data/brick-sdc2/glusterfs/dht2, path creation failed,
reason: No such file or directory
Not good.
Taking out the 03 machine
[root at sfsccl03 ~]# gluster volume create brick2 transport tcp
sfsccl01:/data/glusterfs/dht2 sfsccl02:/data/glusterfs/dht2Creation of
volume brick2 has been successful. Please start the volume to access data.
I am wondering if I should remove the 03 machine from the volume, start
it up with 01 and 02, and then add the 03 machine in, after forcing the
volume back up. Any thoughts?
--
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics Inc.
email: landman at scalableinformatics.com
web : http://scalableinformatics.com
http://scalableinformatics.com/sicluster
phone: +1 734 786 8423 x121
fax : +1 866 888 3112
cell : +1 734 612 4615
Is this normal? Brick3: sfsccl03:/data/brick-sdc2/glusterfs/dht Brick4: sfsccl03:/data/brick-sdd2/glusterfs/dht Both are pointing to the same dir location. Could that be confusing gluster? On Fri, Dec 2, 2011 at 10:24 AM, Joe Landman <landman at scalableinformatics.com> wrote:> Can't start a volume. > > [root at sfsccl03 ~]# gluster volume start brick1 > brick: sfsccl03:/data/brick-sdc2/glusterfs/dht, path creation failed, > reason: No such file or directory > > But ... > > > [root at sfsccl03 ~]# ls -alF /data/brick-sdc2/glusterfs > total 0 > drwxr-xr-x 4 root root ?27 Dec ?2 13:00 ./ > drwxr-xr-x 4 root root 107 Jul ?5 11:55 ../ > drwxrwxrwt 7 root root ?61 Sep 15 11:35 dht/ > drwxr-xr-x 2 root root ? 6 Dec ?2 13:00 dht2/ > > So it is there. > > [root at sfsccl03 ~]# ls -alF /data/brick-sdc2/glusterfs/dht > total 128 > drwxrwxrwt ? ?7 root ? root ? ? ? ?61 Sep 15 11:35 ./ > drwxr-xr-x ? ?4 root ? root ? ? ? ?27 Dec ?2 13:00 ../ > drwxr-xr-x 1230 root ? root ? ? 65536 Oct 24 09:14 equity/ > drwxr-xr-x 1740 oracle root ? ? 65536 Nov 30 23:33 opra/ > drwxr-xr-x ? 35 oracle oinstall ? 501 Jul ?9 17:07 tag/ > drwxr-xr-x ? 11 root ? root ? ? ? 126 Jul ?1 08:51 taq/ > drwxr-xr-x ? ?2 root ? root ? ? ? ?34 Jul 11 19:44 test/ > > > and it is readable. > > More info: > > [root at sfsccl03 ~]# gluster volume info brick1 > > Volume Name: brick1 > Type: Distribute > Status: Stopped > Number of Bricks: 4 > Transport-type: tcp > Bricks: > Brick1: sfsccl01:/data/glusterfs/dht > Brick2: sfsccl02:/data/glusterfs/dht > Brick3: sfsccl03:/data/brick-sdc2/glusterfs/dht > Brick4: sfsccl03:/data/brick-sdd2/glusterfs/dht > > [root at sfsccl03 ~]# gluster peer status > Number of Peers: 2 > > Hostname: sfsccl02 > Uuid: 6e72d1a8-bdeb-4bfb-806c-7fa8b98cb697 > State: Peer in Cluster (Connected) > > Hostname: sfsccl01 > Uuid: 116197cd-5dfe-4881-85ad-5de2be484ba6 > State: Peer in Cluster (Connected) > > a volume reset doesn't help. > > [root at sfsccl03 ~]# gluster volume reset brick1 > reset volume successful > > [root at sfsccl03 ~]# gluster volume start brick1 > brick: sfsccl03:/data/brick-sdc2/glusterfs/dht, path creation failed, > reason: No such file or directory > > New volume creation also fails. > > [root at sfsccl03 ~]# gluster volume create brick2 transport tcp > sfsccl01:/data/glusterfs/dht2 sfsccl03:/data/brick-sdc2/glusterfs/dht2 > sfsccl02:/data/glusterfs/dht2 sfsccl03:/data/brick-sdd2/glusterfs/dht2 > brick: sfsccl03:/data/brick-sdc2/glusterfs/dht2, path creation failed, > reason: No such file or directory > > Not good. > > Taking out the 03 machine > > [root at sfsccl03 ~]# gluster volume create brick2 transport tcp > sfsccl01:/data/glusterfs/dht2 ?sfsccl02:/data/glusterfs/dht2Creation of > volume brick2 has been successful. Please start the volume to access data. > > I am wondering if I should remove the 03 machine from the volume, start it > up with 01 and 02, and then add the 03 machine in, after forcing the volume > back up. ?Any thoughts? > > -- > Joseph Landman, Ph.D > Founder and CEO > Scalable Informatics Inc. > email: landman at scalableinformatics.com > web ?: http://scalableinformatics.com > ? ? ? http://scalableinformatics.com/sicluster > phone: +1 734 786 8423 x121 > fax ?: +1 866 888 3112 > cell : +1 734 612 4615 > _______________________________________________ > Gluster-users mailing list > Gluster-users at gluster.org > http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
On Fri, 02 Dec 2011 13:24:26 -0500 Joe Landman <landman at scalableinformatics.com> wrote:> Can't start a volume. > > [root at sfsccl03 ~]# gluster volume start brick1 > brick: sfsccl03:/data/brick-sdc2/glusterfs/dht, path creation failed, > reason: No such file or directory > > But ... > > > [root at sfsccl03 ~]# ls -alF /data/brick-sdc2/glusterfs > total 0 > drwxr-xr-x 4 root root 27 Dec 2 13:00 ./ > drwxr-xr-x 4 root root 107 Jul 5 11:55 ../ > drwxrwxrwt 7 root root 61 Sep 15 11:35 dht/ > drwxr-xr-x 2 root root 6 Dec 2 13:00 dht2/The mode on /data/brick-sdc2/glusterfs/dht - especially the sticky bit set - seems odd. Have you looked at the xattrs on that?