HACMP: clstat 1

[ogniemi]

Hello,

I've working cascading cluster HACMP build of 2 nodes. Everything is
fine but one thing: "clstat" doesn't discover the cluster status -
shows the following:

clstat - HACMP Cluster Status Monitor
-------------------------------------


THERE ARE NO CLUSTERS CURRENTLY ACTIVE



THE PROGRAM WILL CONTINUE SEARCHING FOR ONE

The clinfo and snmpd is not customized (default community string:
public)

# lssrc -g cluster
Subsystem Group PID Status
clstrmgr cluster 14440 active
clsmuxpd cluster 15532 active
clinfo cluster 17710 active

# clfindres
GroupName Type State Location Sticky Loc
------------- ---------- ------ -------- ----------
res2 cascading UP clnode02
res2_2 cascading UP clnode02
res1 cascading UP clnode01
res1_2 cascading UP clnode01


# lssrc -Ss clinfo
#subsysname:synonym:cmdargsath:uid:auditid:standin:standout:standerr:action:multi:contact:svrkey:svrmtyperiority:signorm:sigforce:display:waittime:grpname:
clinfo:::/usr/sbin/cluster/clinfo:0:0:/dev/console:/dev/console:/dev/console:-O:-Q:-K:0:0:20:0:0:-D:15:cluster:



Please answer what is missed? What should be done to have the real
status of the cluster?


thx in advance,
Mark

 

[DSMARWAY]

hi ,

what version of the operating system & HACMP have you got installed ?

On version 5.2 there is a problem with snmpd version3 , hence when you run clstat it doesn't display the state of the cluster.

what maintenance level, OS and HA have you got installed ?

 

[ogniemi]

thx for reply,

I've HACMP 4.5 AIX 5.2 ML_02.

regards,
Mark

 

[ogniemi]

Does anyone know the solution to it?


thx in advance,
M,

 

[DSMARWAY]

Hi,

firstly go to the latest HACMP patches , i think they might be 4.5.0.7 or higher , if not already done so.

If the problem still exists after the patches have been installed , degrade the snmpd daemon to version one :-
Note:- to do this , i would stop your cluster , and do this on both servers or more .

cd /usr/sbin
ls -l snmpd ( should be link to snmpdv3ne )

Run /usr/sbin/snmpv3_ssw -1

then check the link again it should read snmpdv1


HTH

 

[ogniemi]

Thx, DSMARWAY

I did you suggested and works properly now.

- upgraded HACMP to 4.5.0.10 (didn't help)

- dgradated SNMP v.3 to SNMP v.1 (helpded)

thx,
M.

 

[Hippster]

I have AIX 4330-11 and HACMP 4.2. I made the change to use snmpd Version 1 and snmp works great, however, the clstat still says no active clusters. My output is such:

lssrc -g cluster
Subsystem Group PID Status
clstrmgr cluster 15746 active
clsmuxpd cluster 16026 active
clinfo cluster 16530 active

lssrc -Ss clinfo
#subsysname:synonym:cmdargsath:uid:auditid:standin:standout:standerr:action:multi:contact:svrkey:svrmtyperiority:signorm:sigforce:display:waittime:grpname:
clinfo::-c:/usr/sbin/cluster/clinfo:0:0:/dev/null:/dev/null:/dev/null:-R:-Q:-K:0:0:20:0:0:-D:15:cluster:

clstat - HACMP Cluster Status Monitor
---------------------------------------------


THERE ARE NO CLUSTERS CURRENTLY ACTIVE



THE PROGRAM WILL CONTINUE SEARCHING FOR ONE

Can someone give me some ideas about this... ?

Thanks

 

[Hippster]

Yet another problem....

I am constantly getting these errors in my snmpd.log file.

12/17/03 11:08:08 NOTICE: SMUX relation started with (127.0.0.1+53898+19382)
12/17/03 11:08:08 NOTICE: SMUX packet from (127.0.0.1+53898+19382)
12/17/03 11:08:08 EXCEPTIONS: simpleOpen rejected (badIdentity): risc6000clsmuxpd (SMUX 127.0.0.1+53898+19382)

I need the OID of the clsmuxpd...

Can one of the snmpd experts help me on this..

Thanks,

Don