Oracle process hangs while writing on NFS disk.

[Evgeny Filippenkov]

Hello !
I tried to use NFS disk for oracle backup by rman. But server process hangs after it wrote any file to NFS disk. It doesn't metter what kind of writing I use. It may be rman backup or command "alter database backup controlfile to ..." or utl_file. The result looks similar. File is created but control doesn't return to clients process. However I can read and write any files from shell on this NFS disk and have no problem.
My system is K370 with HP-UX 11.00 and Oracle 8.1.7.4 64 bit.
Does anybody have experience in this problem. I guess some specific options exist for 'right' mounting NFS filesystems for oracle.
Thanks for advance.

[Steven E. Protter]

What version of NFS is being used. If its v2, it could be the 2 GB file size limit.

If its nfs 3, the fileystem itself might not be set of for larrge files.

SEP

[Evgeny Filippenkov]

Thanks. But how can I know what version of NFS is in my system ? BTW I think the problem is not here. I tried to write files 10-50 Mb and didn't get success.

[Steven E. Protter]

nfsstat command displays processes and version nubers.

swlist -i

You can browse down to your NFS software and determine the version numbers.

Somewhere in there are the version numbers. nfsstat, if there is no version 3, that explains the issue, since versions below v3 did not support 2 GB and larger filesystems.

SEP

[Evgeny Filippenkov]

Thanks but I did write the files were less then 2 gigs. Below I put "nfsstat" command result. People say the problem maybe raised because oracle and nfs can't "negotiate" about file locking.

Server rpc:
Connection oriented:
N/A
Connectionless oriented:
calls badcalls nullrecv
0 0 0
badlen xdrcall dupchecks
0 0 0
dupreqs
0

Server nfs:
calls badcalls
0 0
Version 2: (0 calls)
null getattr setattr
0 0% 0 0% 0 0%
root lookup readlink
0 0% 0 0% 0 0%
read wrcache write
0 0% 0 0% 0 0%
create remove rename
0 0% 0 0% 0 0%
link symlink mkdir
0 0% 0 0% 0 0%
rmdir readdir statfs
0 0% 0 0% 0 0%
Version 3: (0 calls)
null getattr setattr
0 0% 0 0% 0 0%
lookup access readlink
0 0% 0 0% 0 0%
read write create
0 0% 0 0% 0 0%
mkdir symlink mknod
0 0% 0 0% 0 0%
remove rmdir rename
0 0% 0 0% 0 0%
link readdir readdir+
0 0% 0 0% 0 0%
fsstat fsinfo pathconf
0 0% 0 0% 0 0%
commit
0 0%

Client rpc:
Connection oriented:
N/A
Connectionless oriented:
calls badcalls retrans
58163 0 0
badxids timeouts waits
0 0 0
newcreds badverfs timers
0 0 5
toobig nomem cantsend
0 0 0
bufulocks
0

Client nfs:
calls badcalls clgets
58163 0 58163
cltoomany
0
Version 2: (0 calls)
null getattr setattr
0 0% 0 0% 0 0%
root lookup readlink
0 0% 0 0% 0 0%
read wrcache write
0 0% 0 0% 0 0%
create remove rename
0 0% 0 0% 0 0%
link symlink mkdir
0 0% 0 0% 0 0%
rmdir readdir statfs
0 0% 0 0% 0 0%
Version 3: (58163 calls)
null getattr setattr
0 0% 160 0% 2 0%
lookup access readlink
19 0% 33 0% 0 0%
read write create
0 0% 57664 99% 1 0%
mkdir symlink mknod
1 0% 0 0% 0 0%
remove rmdir rename
1 0% 0 0% 0 0%
link readdir readdir+
0 0% 0 0% 9 0%
fsstat fsinfo pathconf
68 0% 1 0% 1 0%
commit
203 0%

[Tom Wilson_6]

I'm seeing this exact same problem. Does anyone know what the resolution to this was?

[Brian Crabtree]

Tom,

I would open up a new issue for this, so you have some control over it, but I will bite. I wouldn't use NFS for this, however, have you tried querying the v$session_wait for the sid that you are using and see what it is waiting on.

Looking on metalink, one possible try would be using the "nolock" functionality.

Thanks,

Brian

[Sanjay Kumar Suri]

Have a look at this document from Oracle:

http://otn.oracle.com/deploy/availability/pdf/oscp_nfs_hi_tests.pdf


sks

[T G Manikandan]

First check your database alert log.
That should give valuable info.

Also check your syslog.log at /var/adm/syslog.


Revert on this/.

[RolandH]

You can check what NFS version is available on your system with "rpcinfo -u nfs".

Which version is used you can see if you use the command "nfsstat -m" on your nfs-client system.

Use hard nfs-mounts for your copy. Then you can be sure that all your backup data is copied to you backup-system.

I recommend you the Patch PHNE_29302 for your System, too.

Roland