HACMP & Delayed spooling

[rlake]

Chaps,

Any ideas on this would be a great help:-

AIX 4.3.3, HACMP 4.4.1 ES.

1. HACMP cluster services running - spooling works fine.
2. HACMP cluster services NOT running - spooling takes 3 to 4 minutes?.

I should add that we are printing remotely across the network using the qdaemon.

REALLY dont understand this one - I could maybe have more to work on if printing didn't work at all, but it DOES...

Any ideas would be greatfully accepted..

 

[sbix]

HACMP often uses MAC address overwriting, in some cases some hub-switch have to execute a "sweep cycle" in order to remap MAC address on physical board.
Furthermore, how do the clients resolve the server's IP address?
Is DNS used?
Is the server's boot IP address known to the DNS or it's only into clients "host" files?

 

[rlake]

DNS is NOT used in this instance. Its all done by explicit IP and, as far as I understand it, HACMP in this environment is NOT configured for MAC address takeover.

 

[sbix]

Are these client known in the /etc/hosts ?

 

[rlake]

Yes. Everything is resolveable in /etc/hosts.

 

[sbix]

A telnet session from one of these clients has the same behaviour?
Is it normal in ESTABLISHING while HACMP is on and slow when HACMP is off?

 

[rlake]

Telnet is NOT affected regardless as to whether HACMP is running or not. Only spooling is affected.