Re: [nv-l] Question#1 - Redundant & Disaster Recovery Solution for (AIX) Net view Server

[Bill Evans <wvevans@attglobal.net>]

I've implemented a solution for
a customer where:

• We stop netmon each night at three, TAR up the database and configuration, archive the copy restart netmon and send a custom trap to the backup machine.  It's the easiest way I've found to synchronize the processes.
• On the backup machine the custom trap drives a script to  remote copy the files to the backup machine, archive the backup databases and configuration, stop netmon,  restore databases and configuration with PAX and TAR, do a resetci and mapadmin and restart netmon). 
• We have each machine monitor the network and the other machine.  Polling is set to the normal five minutes on each server  for the other NetView server.  It could be less if you want faster takeover .  Polling is set to the normal five minutes on the primary machine and to 60 minutes on the backup. Loss of either NetView server causes a high priority message to the operations center operators.  
• I wrote a script to switch polling from five to sixty minutes and back.  That procedure is run to make polling the network every five minutes if the secondary system  detects the primary has failed and reset it to sixty minutes when the primary returns.  This is only needed if the customer worries about the bandwidth the status polling is using.  

The only trouble spot I've found is the need to preserve the TMR OID if NetView
is working under the Tivoli Framework.

Bill Evans