Skip to contentSkip to conversation

Stale handle error on RHEL4 (2.6.9-22.ELsmp)

DISCUSSION mehta kiran 2 participants 6 messages Feb 16, 2006 – Feb 18, 2006

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Hi,
When nfs(v4) server fails over from one system to=20 other it gets stale handle error.
All systems have RHEL4 installed (same kernel) uname -a =3D 2.6.9-22.ELsmp #1 SMP=20
=20

Details :

[root@vcslinux118 bin]# mount
<---snip>
sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw) nfsd on /proc/fs/nfsd type nfsd (rw) /dev/sdi2 on /dir2 type ext3 (rw)
/dev/sdi1 on /dir1 type ext3 (rw)
/dir2 on /export/exp2 type bind (rw,bind) /dir1 on /export/exp1 type bind (rw,bind) </---snip>

[root@vcslinux118 bin]# exportfs -v /export/exp1 =20 <world>(rw,wdelay,nohide,no_root_squash) /export/exp2 =20 <world>(rw,wdelay,nohide,no_root_squash)
/export =20
<world>(rw,wdelay,nohide,no_root_squash,fsid=3D0) [root@vcslinux118 bin]# uname -a
Linux vcslinux118.vxindia.veritas.com 2.6.9-22.ELsmp #1 SMP Mon Sep 19 18:00:54 EDT 2005 x86_64 x86_64 x86_64 GNU/Linux

I have mounted this on client as=20
#mount -t nfs4 10.212.102.70:/ /kiran

After failover i get stale handle error on client. Is this a known issue ?

thanks,
kiran
=20

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around=20
http://mail.yahoo.com=20

-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les
for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D=
121642
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

#1 lore raw
Quoted 31 lines · mehta kiran

Hi,
When nfs(v4) server fails over from one system to other it gets stale handle error.
All systems have RHEL4 installed (same kernel) uname -a = 2.6.9-22.ELsmp #1 SMP

Details :

[root@vcslinux118 bin]# mount
<---snip>
sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw) nfsd on /proc/fs/nfsd type nfsd (rw) /dev/sdi2 on /dir2 type ext3 (rw)
/dev/sdi1 on /dir1 type ext3 (rw)
/dir2 on /export/exp2 type bind (rw,bind) /dir1 on /export/exp1 type bind (rw,bind) </---snip>

[root@vcslinux118 bin]# exportfs -v /export/exp1 <world>(rw,wdelay,nohide,no_root_squash) /export/exp2 <world>(rw,wdelay,nohide,no_root_squash) /export <world>(rw,wdelay,nohide,no_root_squash,fsid=0) [root@vcslinux118 bin]# uname -a
Linux vcslinux118.vxindia.veritas.com 2.6.9-22.ELsmp #1 SMP Mon Sep 19 18:00:54 EDT 2005 x86_64 x86_64 x86_64 GNU/Linux

Quoted 5 lines

I have mounted this on client as
#mount -t nfs4 10.212.102.70:/ /kiran

After failover i get stale handle error on client. Is this a known issue ?

It is expected, yes. Your failover system has to be a precise copy of the original filesystem. If the inode number differs for a given file or directory, then the filehandles will differ on the original system and the failover system, and so you will get ESTALE errors.

Cheers,
Trond

-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

#2 lore raw

Hi Trond,
partitions /dev/sdi1 and /dev/sdi2 are=20
seen by both the systems on which server
can start and their major/minor numbers
are same on both the systems.
Sfter server fails over to other system,
output of exportfs and mount looks the same.
Should /export directory which is exported
with fsid=3D0 be on shared disk as well ?
=20
thanks,
kiran
=20
=20

--- Trond Myklebust <[email protected]>

Quoted 41 lines · wrote:
Quoted 2 lines · mehta kiran

Hi,
When nfs(v4) server fails over from one system

to=20

Quoted 10 lines

other it gets stale handle error.
All systems have RHEL4 installed (same kernel) uname -a =3D 2.6.9-22.ELsmp #1 SMP=20
=20
=20
Details :
=20
[root@vcslinux118 bin]# mount
<---snip>
sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs

(rw)

Quoted 8 lines

nfsd on /proc/fs/nfsd type nfsd (rw)
/dev/sdi2 on /dir2 type ext3 (rw)
/dev/sdi1 on /dir1 type ext3 (rw)
/dir2 on /export/exp2 type bind (rw,bind)
/dir1 on /export/exp1 type bind (rw,bind)
</---snip>
=20
[root@vcslinux118 bin]# exportfs -v

Quoted 8 lines

/export/exp1 =20
<world>(rw,wdelay,nohide,no_root_squash)
/export/exp2 =20
<world>(rw,wdelay,nohide,no_root_squash)
/export =20
<world>(rw,wdelay,nohide,no_root_squash,fsid=3D0)
[root@vcslinux118 bin]# uname -a
Linux vcslinux118.vxindia.veritas.com

2.6.9-22.ELsmp

Quoted 8 lines

#1 SMP Mon Sep 19 18:00:54 EDT 2005 x86_64 x86_64 x86_64 GNU/Linux
=20
I have mounted this on client as=20
#mount -t nfs4 10.212.102.70:/ /kiran
=20
After failover i get stale handle error on client. Is this a known issue ?

Quoted 16 lines

=20
It is expected, yes. Your failover system has to be a precise copy of
the original filesystem. If the inode number differs for a given file or
directory, then the filehandles will differ on the original system and
the failover system, and so you will get ESTALE errors.
=20
Cheers,
Trond
=20
=20
=20

-------------------------------------------------------

Quoted 7 lines

This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the=20 web. DOWNLOAD SPLUNK!

http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D=
121642

Quoted 4 lines

_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
=20

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around=20
http://mail.yahoo.com=20

-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les
for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D=
121642
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

#3 lore raw
Quoted 9 lines · mehta kiran

Hi Trond,
partitions /dev/sdi1 and /dev/sdi2 are
seen by both the systems on which server
can start and their major/minor numbers
are same on both the systems.
Sfter server fails over to other system,
output of exportfs and mount looks the same.
Should /export directory which is exported
with fsid=0 be on shared disk as well ?

Probably, yes.

Cheers,
Trond

-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

#4 lore raw

Hi Trond,
But this is different than behaviour of NFS=20 on other platforms. On other platforms like=20 AIX, root of exported fses need not be=20 on shared disk.
Anyways, i will try out the thing suggested by you and post the result tomorrow.

thanks,
kiran

--- Trond Myklebust <[email protected]>

Quoted 11 lines · wrote:
Quoted 9 lines · mehta kiran

Hi Trond,
partitions /dev/sdi1 and /dev/sdi2 are=20
seen by both the systems on which server
can start and their major/minor numbers
are same on both the systems.
Sfter server fails over to other system,
output of exportfs and mount looks the same.
Should /export directory which is exported
with fsid=3D0 be on shared disk as well ?

Quoted 7 lines

=20
Probably, yes.
=20
Cheers,
Trond
=20
=20

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around=20
http://mail.yahoo.com=20

-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les
for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D=
121642
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

#5 lore raw

Hi Trond,
I tried with a setup in which filesystem exported with fsid=3D0 is also on shared disk. I worked that way. However it is still interesting to see that on AIX it works with a little less restriction as mentioned in my mail below.
=20
thanks,
kiran

     =20
Quoted 33 lines · --- mehta kiran

Hi Trond,
But this is different than behaviour of NFS=20 on other platforms. On other platforms like=20 AIX, root of exported fses need not be=20 on shared disk.
Anyways, i will try out the thing suggested by you and post the result tomorrow.
=20
thanks,
kiran
=20
=20
--- Trond Myklebust <[email protected]> wrote:
=20

Quoted 11 lines
Quoted 9 lines · mehta kiran

Hi Trond,
partitions /dev/sdi1 and /dev/sdi2 are=20
seen by both the systems on which server
can start and their major/minor numbers
are same on both the systems.
Sfter server fails over to other system,
output of exportfs and mount looks the same.
Should /export directory which is exported
with fsid=3D0 be on shared disk as well ?

Quoted 7 lines

=20
Probably, yes.
=20
Cheers,
Trond
=20
=20

Quoted 10 lines

=20
=20
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam
protection around=20
http://mail.yahoo.com=20
=20
=20

-------------------------------------------------------

Quoted 7 lines

This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the=20 web. DOWNLOAD SPLUNK!

http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D=
121642

Quoted 4 lines

_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
=20

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around=20
http://mail.yahoo.com=20

-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les
for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D=
121642
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs