I’ve been doing this for a long time … but when I read the install guide for Bb Vista 4.2, my interpretive filters must’ve been clogged. I could’ve sworn it wanted me to use a separate script to upgrade the 3rd node of our small three node cluster. We have two node development and preproduction clusters, so I had no environment to practice the upgrade on.
Stupid. Stupid. Stupid. Eventually I figured out you only need to use the serverConfig.sh if you are ADDING a node over 2. Node 3 or Node 53 would be installed with that script. But not upgraded. If the cluster already knows about the node, if it’s an upgrade, this script is superfluous.
The documentation has tended to be unclear on its use of the language Upgrade vs Install either … It also tends to confuse actions done on the Admin node versus Managed Nodes.
It’s a good thing we’re professionals here. Many claim we don’t read at all. For pete’s sake, we claim we don’t read at all. RTFM we tell each other.
If this stuff from Bb is an example of the reading material I have to look forward to, is it any wonder we just make it up as we go? Figure it out ourselves?
I know. I should’ve asked someone.
Oh well. I installed the 3rd node today the regular way. Three months after the fact. I did an application only, upgrade, and I slapped AP2 and a Hotfix on without restarting in between since the rest of my cluster was at that level. Of course you also have to know all the undocumented features like the size of the database field for the Admin DNS name, so put the IP address in startup.properties instead if your DNS name is overly long; and find a copy of the missing wlstProfile.py , slap that into your /WebCTDomain. Don’t forget your SSL cert paths get wiped out of your Chat-Serverconfig.xml. Oh, and the classpath for the Weblogic patches in startup.properties needs changing too; still refers to weblogic 8.