Host name problem.,,,,,,,

[Kannandgl_1]

Greetings Friends ,

Hi , i am getting small problem in one of the host.
when ever i am trying to get the host name its showing original name .At the same time i am using uname -a command i am getting below output.Please any one advise to me .. where is node name is taken . i am checking the /etc/rc.config.d/netconf file also ,their correct entry .
=======================================
#uname -a

HP-UX unknown B.11.23 U ia64 4257359972 unlimited-user license
=======================================

[Asif Sharif · Accepted solution]

Hi,

see the below link.

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1145232

Regards,
Asif Sharif

[Raj D.]

Kannandgl,
Check and make sure you have no junk files under /etc/rc.config.d/
- You are not supposed to keep any backup file under /etc/rc.config.d/ , like netconf.old ore netconf.08092010 etc , and it may cause the problem.
- If there is any such file move it and reboot and you can check, most of the time that causes the problem.

Hth,
Raj.

[Raj D.]

Also check with this with debug mode & see what you are getting,

sh -x /sbin/rc1.d/S320hostname

hth,
Raj.

[Kannandgl_1]

Dear RAJ,

1.Today i am coping the file same location ..
2.I am posting the Output # sh -x /sbin/rc1.d/S320hostname.

========================
+ PATH=/sbin:/usr/sbin:/usr/bin
+ export PATH
+ rval=0
+ echo usage: /sbin/rc1.d/S320hostname {start}
usage: /sbin/rc1.d/S320hostname {start}
+ exit 0
========================

[Raj D.]

Kannan,

Try,

sh -x /sbin/rc1.d/S320hostname start

[Raj D.]

I think you can try with stop and start, and see what happens, this script executes in run level 1, so there may be some dependency ,

- Do you see any error in /etc/rc.log
# cat /etc/rc.log | grep -i -e err -e fail

see in /etc/rc.log for the below portion:


Setting hostname
Output from "/sbin/rc1.d/S320hostname start":
----------------------------



- since when you are getting the hostname as unknown, what changed since last time it was working fine.


- check netconf file again.
# cat /etc/rc.config.d/netconf | grep -v ^#

hth,
Raj.

[Kannandgl_1]

Dear RAJ,

+ hostname hogisdata2
+ set_return
+ [ -z hogisdata2 ]
+ uname -S hogisdata2
Nodename must be less than 9 characters
+ set_return
EXIT CODE: 1
+ [ -z HP-UX ]
+ setuname -t -s HP-UX
+ set_return
+ exit 1

Regards
Kanna

[Kannandgl_1]

Greetings Raj,

As per your valuable answer, i got the faile error on this node.
========================
cat /etc/rc.log | grep -i -e err -e fail
/sbin/rc1.d/S320hostname start" FAILED
rw-r--r-- 1 root sys 0 Sep 3 2007 swspluninfo.error
=========================

Regards
Kanna

[Raj D.]

Kannan,

you have set a wrong host name, you have to limit it to 8 character:

>>
+ uname -S hogisdata2
Nodename must be less than 9 characters
>>



- Change to 8 character hostname:
1. Correct /etc/rc.config.d/netconf entry.
2. Correct /etc/hosts entry.
3. reboot and check.


or
- /sbin/set_parms hostname
And type 8 character hostname, and reboot.

This should fix the issue.



Hth,
Raj.

[Raj D.]

And after the reboot,
again check: with the below command , it should not be failed once it is satisfied with 8 character correct hostname,

>>
cat /etc/rc.log | grep -i -e err -e fail
/sbin/rc1.d/S320hostname start" FAILED
>>
- Once the hostname script done in the run level 1, the "S320hostname start" should be ok with no failure message in rc.log .



- And finally verify, that the host name is correct in all the places:
# uname -a ; hostname
# cat /etc/rc.config.d/netconf | grep -v ^#

# cat /etc/hosts | grep `hostname`

# ping `hostname`
# nslookup `hostname`



Hth,
Raj.

[Kannandgl_1]

Dear Raj,

Thanks for your support.I will try to change the host name.But the problem is we are using 2 node cluster.If any other alternative ways is there.I have to changed both nodes and reconfigured the cluster file that way i worries jeee...

Kindly give the some step or idea for me reconfigured the 2 node cluster.

Thanks and regards
Kanna

[Raj D.]

Kanna,

Check this out for MC/SG cluster configuration:

- You can start with , first fix the hostname issues, correct ip adress /etc/hosts etc.
- put each host name entrt in /etc/cmcluster/cmclnodelist
- Then start with:


# swlist -l product | grep -i servicegu
# cmquerycl -v -C /etc/cmcluster/cluster1.ascii -n node1 -n node2 -q quorumserver

- Modify the cluster.ascii file per the config.
- next cmcheckconf/cmapplyconf/ --> cmrunnode -> if everything is ok proceed to package configuration. --> after that, failover test. verify --> Done.


- Further go through:
http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1224594


- For diffrent topic , remember to open a different ITRC thread.


Cheers,
Raj.

[Asif Sharif]

Changing HOSTNAME on a ServiceGuard node

See the Technical knowledge base - document
UMCSGKBRC00006541
http://www12.itrc.hp.com/service/cki/docDisplay.do?docLocale=en&docId=emr_na-c01028186-2

http://forums13.itrc.hp.com/service/forums/questionanswer.do?admit=109447627+1281499722296+28353475&threadId=1266876


Regards,
Asif Sharif