Can't join node to cluster

[Pattarapong Noiruang]

After I install patch PHSS_30723. I found some problem. First, I can't join that node to cluster. After I run 'cmrunnode -v msmp1', it's not success. I found some messages in syslog file:
Sep 6 10:00:50 msmp1 cmclconfd[8646]: FtoM - flat header info: magic_number=670
918047 format_version=7 data_version=1 length=15248 size=15248 crc=28753
Sep 6 10:00:50 msmp1 cmclconfd[8646]: Local node msmp1 (B.11.00,11.09) is not r
unning in cluster clsmp (1088535314,1) as node 1
Sep 6 10:00:50 msmp1 cmclconfd[8646]: Unable to connect to 127.0.0.1,5304: 239
Sep 6 10:00:50 msmp1 above message repeats 2 times
Sep 6 10:00:50 msmp1 cmclconfd[8646]: Waiting for next TCP message
Sep 6 10:00:50 msmp1 cmclconfd[8646]: Stopped receiving TCP messages (Software
caused connection abort)

Another problem: I found more problems, example,
-passwd file is changed(for some user).
-can't enter some menu in SAM
-can't use man command, it will return :
msmp1:/var/adm/syslog#man cmrunnode
man: Syntax error at line 1 : `(' is not expected.

Pls. feed back me 2 ways by email and this forrum. I can't online alltime

Thank You

[Sridhar Bhaskarla]

Hi,

Was the patch install successful?. Try

swlist -l patch -a state PHSS_30723

It should say configured|applied.

If not, then there might be some issues you may find in swagent.log file that you have to fix.

If the patch install was successful, then I suggest you try the following on msp1 itself.

1. Do

inetd -l (if -l is not already enabled)

remsh msp1 -l root

It should be successful if you are using .rhosts. Otherwise, look at /var/adm/syslog/syslog.log and see how msp1 is appearing as. It should list correct hostname/IP address.

2. Make sure you have all hacl* ports open on the system. 'grep hacl /etc/services' and the corresponding entries in /etc/inetd.conf. Ensure you are not blocking access to port 5304/hacl-cfg in /var/adm/inetd.sec

Your attached syslog.log doesn't show many details.

-Sri