pmd lost contact with ipmap

[DTMan]

After reviewing our logs this morning I noticed the following errored posted twice around 0223 and 0357. It also says that pid ungracefully disconnected from pmd with 0 pending events. Error on socket #46 src:ipmap pid 4554 and 5407. Lost connection with pmd/ovevent process (application disconnected). Any ideas?

DTMan

 

[vlan52]

Could you tell us which versions are you currently running for NNM/VPO/HP-OV?
OS?

Thanks,
vlan52
vlan52
The end of wisdom is freedom. The end of culture is perfection. The end of
education is character. The end of knowledge is love.

 

[DTMan]

Yes, we are running a Solaris 2.6 box with HP OV 6.10.

 

[vlan52]

This could a problem with one of the traps received by pmd. Or that pmd queues are not large enough try to change the maximum number of events which will be queued for an application to 2048, change the pmd.lrf file to the following q2048 eg:

OVs_YES_START::-SOV_EVENT;q2048:OVs_WELL_BEHAVED:15AUSE

Then ovaddobj pmd.lrf and restart NNM.

Verify as well is that your name resolver (DNS) is working fine since a bad DNS will affect pmd.

To verify your DNS .

Find the checkDNS.ovpl tool under /opt/OV/support/checkDNS.ovpl

At the command promI pt type:
checkDNS.ovpl -v

Every device listed in the NNM topology database is issued an appropriate gethostbyaddr or gethostbyname system call.
Addresses or hostnames that are too slow in responding are identified in the output on the screen. Clean up the name resolution for
addresses or hostnames identified as being too slow.


HTH
vlan52
vlan52
The end of wisdom is freedom. The end of culture is perfection. The end of
education is character. The end of knowledge is love.