Latvia Lotus User Group (www.vlaad.lv)

Traveler before 9.0.0.1 is not failing over to other mail cluster node, if primary server is down, and how to avoid it

Vladislav Tatarincev  7 July 2014 13:29:47
If you have two mail servers in cluster and Traveler on separate machine(DMZ), and your traveler below 9.0.1 you may face Traveler outage if Primary server goes down.
"Tell Traveler showuser username" shows user is fine and traveler is aware of second server, but new mails may not arrive because of bug

http://www-01.ibm.com/support/docview.wss?uid=swg1LO78404
This fix will be included in Notes Traveler 9.0.1 IF3 and 9.0.0.1 IF4 and any future releases.  See this technote for the latest maintenance release information.

When you bring primary server down, dont forget to restart Traveler, or you can outomate this by events4.nsf, by creating Domino server monitor.
Selecting that Traveler is monitoring Primary server, and in case server is down for while send a console command "Restart Task Traveler"

default timeout is 3 seconds, I think it is better rise it to 20-30 seconds, to avoid annecessary restarts of Traveler cause by network issues.
Image:Traveler before 9.0.0.1 is not failing over to other mail cluster node, if primary server is down, and how to avoid itImage:Traveler before 9.0.0.1 is not failing over to other mail cluster node, if primary server is down, and how to avoid it
Comments

1Paul Farris  07.07.2014 15:07:18  Traveler before 9.0.0.1 is not failing over to other mail cluster node, if primary server is down, and how to avoid it

Good to know. Thanks, but the command should be "Tell Traveler user username".