NFS is not runing after the kernel tuning

[prasathc]

Hi Admins
I have a problem like i have box {C3000) and I installed 11.11 MCOE with the following bundles.
# swlist
# Initializing...
# Contacting target "test49"...
#
# Target: test49:/
#

#
# Bundle(s):
#

B3899BA B.11.11.01 HP C/ANSI C Developer's Bundle for HP-UX 11.00 (S700)
B3911DB C.03.26 HP aC++ Compiler (S700)
BUNDLE11i B.11.11.0102.2 Required Patch Bundle for HP-UX 11i, February 2001
CDE-English B.11.11 English CDE Environment
FDDI-00 B.11.11.01 PCI FDDI;Supptd HW=A3739A/A3739B;SW=J3626AA
FibrChanl-00 B.11.11.03 PCI FibreChannel;Supptd HW=A5158A
GOLDAPPS11i B.11.11.0312.4 Gold Applications Patches for HP-UX 11i v1, December 2003
GOLDBASE11i B.11.11.0312.4 Gold Base Patches for HP-UX 11i v1, December 2003
GigEther-00 B.11.11.14 PCI/HSC GigEther;Supptd HW=A4926A/A4929A/A4924A/A4925A;SW=J1642AA
HPUXBase64 B.11.11 HP-UX 64-bit Base OS
HPUXBaseAux B.11.11 HP-UX Base OS Auxiliary
HWEnable11i B.11.11.0312.4 Hardware Enablement Patches for HP-UX 11i v1, December 2003
OnlineDiag B.11.11.00.04 HPUX 11.11 Support Tools Bundle
RAID-00 B.11.11.00 PCI RAID; Supptd HW=A5856A
#
# Product(s) not contained in a Bundle:
#

CCASE-MVFS ClearCase
#

After installation of everything (Gold qpk, HWEnable11i) i have been asked to tune some kernel parameters (which is attached) for Clear Case application. After change of kernel parameters i am not able to mount or use any of the NFS resources of box. I explorer the possibilities then i checked that nfs service is not running.

So i did the following
/sbin/init.d/nfs.core stop
/sbin/init.d/nfs.client stop
/sbin/init.d/nfs.server stop

/sbin/init.d/nfs.core start
/sbin/init.d/nfs.client start
/sbin/init.d/nfs.server start

but still i am facing the problem. When i tried to boot with previous kernel (/stand/vmunix_before_kernel_mod {before the modification of kernel paramaters}) the nfs server is running and i am able to mount or use the nfs resources.

I have attached /var/adm/syslog/syslog.log along with this mail.

Can anybody tell me how to handle this situation?

Thanks in Advance
Prasath C

[Stefan Farrelly]

Wheres the parameters they asked you to tune ? not in your attachment.

[prasathc]

Hi
Sorry i forget to attachement.

Here is the attachement.

Thanks in Advance
Prasath C

[Todd McDaniel_1]

It seems odd to me that the kernel parms would do it... I think it is related to the MCOE upgrade... You get the difference with NFS b/c you reverted to a Pre-MCOE Kernel setup... Does the MCOE start the RPC daemons?

Did you check to see if your rpc services are running? I believe you need to have this running on all clients...

You may not have all these running but you should have some rpc procs running.

# pp rpc
root 1256 1 0 Dec 20 ? 0:00 /usr/sbin/rpc.lockd
root 1215 1 0 Dec 20 ? 0:00 /usr/sbin/rpcbind
root 1250 1 0 Dec 20 ? 0:00 /usr/sbin/rpc.statd
root 1651 1 0 Dec 20 ? 2:50 /opt/dce/sbin/rpcd
root 25937 1286 0 Dec 20 ? 0:00 /usr/dt/bin/rpc.ttdbserver


I would double check your /etc/rc.config.d/nfsconf as well just in case.

[Stefan Farrelly]

The new attachment seems to be a master list of all your parameters - it would be far easier to diagnose exactly what changes you were asked to make (compared to the old kernel which works fine).

[Robert-Jan Goossens]

Original question.

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=363346

Prasath,

could you give us the commands you have used and a short list off the kernel parameters you have changed, and how much memory have you got in the c3000 ?

Kind regards,
Robert-Jan

[prasathc]

Hi Robert
Initially i have tried to change all the parameter (which the list i have published already in my previous posting) but the nfs is not working so i thought to tune some of the application related parameters alone (not the full) which is given below.

STRMSGSZ 65535
bufcache_hash_locks 256
create_fastlinks 1
dbc_max_pct 80
default_disk_ir 1
dnlc_hash_locks 512
eqmemsize 2000
maxdsiz 0X20000000
maxdsiz_64bit 0X200000000
maxfiles 1024
maxfiles_lim 2048
maxssiz 0X4000000
maxssiz_64bit 0X4000000
maxswapchunks 4096
maxtsiz 0X40000000
maxtsiz_64bit 0X400000000
maxusers 2048
msgmni 100
msgseg 7168
msgtql 256
ncdnode 300
nfile 30724
nflocks 10000
ninode 19012
nkthread 28723
nproc 16404
npty 256
nstrpty 256
semmni 128
semmns 256
shmmax 0XF000000
st_ats_enabled 1

But still i am facing the nfs problem. Finally i changed only "ninode, nproc, maxuproc, maxswapchnk" parameter and rebuild the kernel, rebooted but no use same NFS problem.
[Note:- All the tuning related activity i am using SAM not with commands.]

Thanks
Prasath c

[prasathc]

Hi

Memory ==> 256 MB
Model ==> 9000/785/C3000

Thanks
Prasath C

[Stefan Farrelly]

Youve really got very little memory on your C-class workstation, 256MB is tiny, especially if youre changing kernel parameters. Just use the /stand/system file which is good (you know if you build a kernel from it and reboot nfs still works) and only change 1 parameter, say maxuprc, then see if nfs it still working.

[Robert-Jan Goossens]

Agree with Stefan, you will need to add more memory, Clear Case likes memory.

Kind regards,
Robert-Jan

[Todd McDaniel_1]

Prasath,

This may be naive on my part, but I think you are going in the wrong direction...

I just dont believe a kernel parm setting can disable your NFS... Id have to see it to believe it.

I would tend to look at Firmware or the OS revision as a possible culprit...

[Robert-Jan Goossens]

Hi Todd,

He is not only getting nfs errors, also

WARNING: modpath: DLKM is not initialized

http://www5.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000065678592

Document description: SYS ADM: "WARNING: modpath: DLKM is not initialized" error
Document id: A5577671

if he boots from his old kernel, all is well.

Regards
Robert-Jan

[prasathc]

Hi All

fine I'll try to change another machine with good memory (if possible). Or else i'll try to upgrade memory in that machine.

About the firmware i am not sure.

And for OS i'll try to get latest OS depot.

Thanks for ur response.
Prasath C