We''re experiencing a very sporadic issue where the client and MDS
appear to be out of sync for some time. See the example below. When
this occurs there are no errors on the mds or any of the oss''s, but
errors on the client like:
Jun 14 13:37:14 bulldogi kernel: Lustre:
3374:0:(namei.c:235:ll_mdc_blocking_ast()) More than 1 alias dir
95854042 alias 2
Jun 14 13:37:14 bulldogi kernel: Lustre:
3374:0:(namei.c:235:ll_mdc_blocking_ast()) Skipped 30 previous similar
messages
Jun 14 13:38:36 bulldogi kernel: Lustre:
3374:0:(namei.c:235:ll_mdc_blocking_ast()) More than 1 alias dir
95854042 alias 2
What does this mean? I see a similar unanswered question on the
mailing list dated May 15th regarding this.
Lustre is not heavily used yet, healthy, and all devices are up.
This problem persists for about one minute before reporting correct information.
Red Hat Enterprise Linux AS release 4
2.6.9-42.0.10.EL_lustre-1.6.0.1smp
[mike@fire kent]$ pwd
/lustre0/njc2/bioInformatics/kent
[mike@fire kent]$ mv src/bin .
[mike@fire kent]$ ls
bin build CVS cvsTest java src utils
[mike@fire kent]$ ls bin
ls: bin: No such file or directory
[mike@fire kent]$ ls -ltr
total 24
?--------- ? ? ? ? ? bin
drwxrwxr-x 5 njc2 njc2 4096 Jun 14 13:22 java
drwxrwxr-x 3 njc2 njc2 4096 Jun 14 13:22 cvsTest
drwxrwxr-x 3 njc2 njc2 4096 Jun 14 13:22 build
drwxrwxr-x 4 njc2 njc2 4096 Jun 14 13:30 utils
drwxrwxr-x 2 njc2 njc2 4096 Jun 14 13:33 CVS
drwxrwxr-x 45 njc2 njc2 4096 Jun 14 13:48 src
Thank You,
jeff
--
Jeff Blasius / jeff.blasius@yale.edu
Phone: (203)432-9940 51 Prospect Rm. 011
High Performance Computing (HPC)
UNIX Systems Administrator, WorkStation Support (WSS)
Yale University Information Technology Services (ITS)
The problem has been fixed via bug 12123, you should see it when 1.6.1 is released. Thanks WangDi Jeff Blasius wrote:> We''re experiencing a very sporadic issue where the client and MDS > appear to be out of sync for some time. See the example below. When > this occurs there are no errors on the mds or any of the oss''s, but > errors on the client like: > Jun 14 13:37:14 bulldogi kernel: Lustre: > 3374:0:(namei.c:235:ll_mdc_blocking_ast()) More than 1 alias dir > 95854042 alias 2 > Jun 14 13:37:14 bulldogi kernel: Lustre: > 3374:0:(namei.c:235:ll_mdc_blocking_ast()) Skipped 30 previous similar > messages > Jun 14 13:38:36 bulldogi kernel: Lustre: > 3374:0:(namei.c:235:ll_mdc_blocking_ast()) More than 1 alias dir > 95854042 alias 2 > > What does this mean? I see a similar unanswered question on the > mailing list dated May 15th regarding this. > > Lustre is not heavily used yet, healthy, and all devices are up. > > This problem persists for about one minute before reporting correct > information. > > Red Hat Enterprise Linux AS release 4 > > 2.6.9-42.0.10.EL_lustre-1.6.0.1smp > > [mike@fire kent]$ pwd > /lustre0/njc2/bioInformatics/kent > [mike@fire kent]$ mv src/bin . > [mike@fire kent]$ ls > bin build CVS cvsTest java src utils > [mike@fire kent]$ ls bin > ls: bin: No such file or directory > [mike@fire kent]$ ls -ltr > total 24 > ?--------- ? ? ? ? ? bin > drwxrwxr-x 5 njc2 njc2 4096 Jun 14 13:22 java > drwxrwxr-x 3 njc2 njc2 4096 Jun 14 13:22 cvsTest > drwxrwxr-x 3 njc2 njc2 4096 Jun 14 13:22 build > drwxrwxr-x 4 njc2 njc2 4096 Jun 14 13:30 utils > drwxrwxr-x 2 njc2 njc2 4096 Jun 14 13:33 CVS > drwxrwxr-x 45 njc2 njc2 4096 Jun 14 13:48 src > > Thank You, > jeff >