NFS mount problem on HP Cluster from linux NFS server.

[jerrym]

I have HP Cluster 11.31 client and Linux SUSE 11 NFS server.

Using /etc/auto.direct and restarting /sbin/init.d/autofs on client then cd into directory I get long wait

then doing ls first time I get after long wait:

NFS readdirplus failed for server panwppnpd: error 5 (RPC: Timed out)

Second ls shows files. As long as it is mounted I can see the files. If automounter unmounts it

and I try to again, I get same problem.

This problem is only with HP Clusters. Other non HP Clusters mount the same export from Linux NFS server okay.

Have not found anything to resolve this problem.

P.S. This thread has been moved from HP-UX>System Administration to HP-UX > networking. -HP Forum Moderator

[Dave Olker]

Posting the AutoFS maps would help.  Do you get the same behavior when you manually mount the exact same filesystem from the exact same server using the exact same options?

Dave

[jerrym]

Linux panwppnpd 3.0.58-0.6.6-default NFS server /etc/exports file entry:

/usr/sap/trans *(rw,sync,no_root_squash,no_subtree_check)

# exportfs

/usr/sap/trans  <world>

HP Cluster /etc/auto.direct:

/usr/sap/trans  panwppnpd:/usr/sap/trans

This only happens when using automounter on HP clusters.

If I manually mount it. It is okay.

[Dave Olker]

Please mount the filesystem manually on the HP-UX client and then post the output of "nfsstat -m" so I can see which mount options are used.

Dave

[jerrym]

#nfsstat -m
/usr/sap/trans from panwppnpd:/usr/sap/trans
 Flags:         vers=3,proto=tcp,sec=sys,hard,intr,link,symlink,acl,devs,rsize=32768,wsize=32768,retrans=5,timeo=600
 Attr cache:    acregmin=3,acregmax=60,acdirmin=30,acdirmax=60

/sapmnt/GCP from moss:/export/sapmnt/GCP
 Flags:         vers=3,proto=udp,sec=sys,soft,intr,link,symlink,acl,nodevs,rsize=32768,wsize=32768,retrans=5,timeo=11
 Attr cache:    acregmin=3,acregmax=60,acdirmin=30,acdirmax=60
 Lookups:     srtt=7 (17ms), dev=3 (15ms), cur=2 (40ms)
 Reads:     srtt=7 (17ms), dev=4 (20ms), cur=2 (40ms)
 Writes:     srtt=7 (17ms), dev=3 (15ms), cur=2 (40ms)
 All:     srtt=7 (17ms), dev=3 (15ms), cur=2 (40ms)

[Dave Olker]

There are two filesystems. Which one is the problem?

[jerrym]

/usr/sap/trans from panwppnpd:/usr/sap/trans
 Flags:         vers=3,proto=tcp,sec=sys,hard,intr,link,symlink,acl,devs,rsize=32768,wsize=32768,retrans=5,timeo=600
 Attr cache:    acregmin=3,acregmax=60,acdirmin=30,acdirmax=60

[Dave Olker]

In your auto.direct map you are not specifying any mount options. What happens if you modify your auto.direct map to specify version 3? The symptom you're describing sounds like an NFS v4 mount point. If you hardcode v3 into the AutoFS map do you get the desired behavior?

Dave

[jerrym]

Its already vers=3 from above.

I remounted with automounter again and here is the nfsstat.

/usr/sap/trans from panwppnpd:/usr/sap/trans
 Flags:         vers=3,proto=udp,sec=sys,soft,intr,link,symlink,acl,nodevs,rsize=32768,wsize=32768,retrans=5,timeo=11
 Attr cache:    acregmin=3,acregmax=60,acdirmin=30,acdirmax=60
 Lookups:     srtt=50 (125ms), dev=5 (25ms), cur=8 (160ms)
 Reads:     srtt=8 (20ms), dev=8 (40ms), cur=5 (100ms)