process "rptdrv' cannot be killed

[Shukor]

Hi there,

Our Backup admin need our help to kill this process "rptdrv" in hpux 11.11 that was chaos their netbackup application. however I failed to kill this process even kill -9.
root@s53kj213[/] # ps -ef |grep -i rptdrv
root 7036 1 0 17:59:46 ? 0:00 bptm -rptdrv -jobid -1255460476 -jm
root 16397 1 0 18:19:46 ? 0:00 bptm -rptdrv -jobid -1255460722 -jm
root 11485 1 0 18:09:46 ? 0:00 bptm -rptdrv -jobid -1255460602 -jm
root 19872 17513 1 18:27:03 pts/td 0:00 grep -i rptdrv
root 22781 1 0 17:29:46 ? 0:00 bptm -rptdrv -jobid -1255460138 -jm
root 27473 1 0 17:39:46 ? 0:00 bptm -rptdrv -jobid -1255460250 -jm

need help please. TQ in advance.

[Mouad_1]

Just a suggestion :
Try kill -24 before (kill -9) as those cannot be trapped :
0 Check access to pid
1 Terminate; can be trapped
2 Terminate; can be trapped
3 Terminate with core dump; can be trapped
9 Forced termination; cannot be trapped
15 Terminate; can be trapped
24 Pause the process; cannot be trapped
25 Pause the process; can be trapped
26 Run a stopped process

[VK2COT]

Hello,

Hmm, the parent process of those bptm processes
is process 1 (init).

You might try to stop all NetBackup
daemons.

I am afraid you might be forced to reboot the
server in the end.

Cheers,

VK2COT

[Shukor]

hi,

already reboot the server but the process "rptdrv" still exist and running. been told by backup admin this process must be killed before netbackup can run normally..FYI, we performed SAN switch and data migration just before this problem occurred..

[Dennis Handly]

>already reboot the server but the process "rptdrv" still exist and running.

It doesn't "still" exist, it was restarted on the reboot. You need to find out who is starting it. I don't see any rptdrv process, just bptm with that parm. So who is starting that process, and why so many?

>Mouad: Try kill -24 before (kill -9) as those cannot be trapped

If kill -9 doesn't work, then -24 won't either.

[VK2COT]

Hello,

These bptm processes cannot happen by magic :)

Somebody or someone started them.

bptm is a NetBackup tape media manager on
master/slave servers and is used when the
storage unit type is Media Manager.

bptm manages the transfer of images between
the client and storage devices. It also
handles communication between backup and
Media Manager software.

Additionally, bptm manages the Netbackup
media database and provides information for
the media list report screen.

bptm is normally started by bpbrm for each
backup/restore operation.

bptm is stopped by a completion of the
operation.

Why don't you check the activity log:

/usr/openv/netbackup/logs/bptm/bptm.log
or
/opt/openv/netbackup/logs/bptm/bptm.log

... on the client machine?

As an alternative, you could simply DISABLE
all NetBackup daemons from starting and
reboot the server. Then see what is going
wrong.

Cheers,

VK2COT

[Shukor]

Hi experts again,

this is the same pending issue but I got some hints from netbackup team. They've claimed that the rpc services is not running proper on my server. They wanted me to verify rpc communication between the my server and the acsls host using the rpcinfo command (rpcinfo -t 300031 1 h). so i ran it but failed;

root@s53kj213[/] # rpcinfo -t acs 30031 1 h
Usage: rpcinfo [-m | -s] [host]
rpcinfo -p [host]
rpcinfo -T netid host prognum [versnum]
rpcinfo -l host prognum versnum
rpcinfo [-n portnum] -u | -t host prognum [versnum]
rpcinfo -a serv_address -T netid prognum [version]
rpcinfo -b prognum versnum
rpcinfo -d [-T netid] prognum versnum

Please advise..how to verify my rpc services?

Thanks.

[Dennis Handly]

>(rpcinfo -t 300031 1 h). I ran it but failed:
># rpcinfo -t acs 30031 1 h

It looks like you leave off the "h".
http://docs.hp.com/en/B2355-60130/rpcinfo.1M.html

[Shukor]

Hi Dennis,

root@s53kj213[/] # rpcinfo -t acs 30031 1
rpcinfo: RPC: Program not registered
program 30031 version 1 is not available

Is the rpv service not running on my server and caused the problem of netbackup to running proper as usual?

[Dennis Handly]

>Is the RPC service not running on my server and caused the problem of netbackup to running proper as usual?

Try "rpcinfo -s" and "rpcinfo -s acs".

[Shukor]

Hi Dennis,

this is the output;

root@s53kj213[/] # rpcinfo -s acs
program version(s) netid(s) service owner
100000 2 tcp,udp rpcbind unknown
536871166 2,1 tcp,udp - unknown
300031 2,1 tcp,udp - unknown

root@s53kj213[/] # rpcinfo -s
program version(s) netid(s) service owner
100000 2,3,4 udp,tcp,ticlts,ticotsord,ticots rpcbind superuser
100024 1 udp,tcp status superuser
100021 2,4,3,1 udp,tcp nlockmgr superuser
100020 1 tcp,udp llockmgr superuser
100068 5,4,3,2 udp cmsd superuser
100083 1 tcp ttdbserver superuser
100005 3,1 tcp,udp mountd superuser
100003 3,2 udp,tcp nfs superuser


but this is the one netbackup team claimed should exist and running;

1073741824 2 tcp 0.0.0.0.194.241 - superuser
1073741825 2 tcp 0.0.0.0.194.252 - superuser

So what the conclusion should i give to them?really appreciates your help..

[Dennis Handly]

>So what the conclusion should I give to them?

Give them the above output and let them make their conclusions? I was just changing their rpcinfo command to something more valid.

[mvpel]

Make sure that what you think is a process refusing to die even with kill -9 is not actually a process being restarted if it's missing. Compare the process ID numbers before and after the kill command, and if they don't match, then the process was, in fact, killed, only it was restarted immediately afterward.

It's pretty rare that a process won't respond to a "kill -9" - the only time I've seen it is when the hung process is in a Processor Set [ https://h20392.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=PSETS ] which has its IOINTR attribute turned off, and the only remedy there is to reboot.

[Dennis Handly]

>mvpel: It's pretty rare that a process won't respond to a "kill -9"

By the time it's in the ITRC forum, we have weeded out the all but "kill -9" and zombie cases. :-)