IBM's ifix/iFeature strategy
Accepted answer
The iFix CAN be installed BEFORE the upgrade.
For 5.0.2 the iFix SHOULD be installed BEFORE the upgrade, because it contains a fix for the upgrade itself. (See http://www-01.ibm.com/support/docview.wss?uid=swg21695623 and defect 343736 .)
Ruby
Martha (Ruby) Andrews
Jazz Foundation L3 Development Team Lead
Comments
The "application/server" is down before/during the upgrade. If I'm correct, the patch is "installed" when the server is up and running. So..., creating the directory and copying the patch...does the system "read" this for the upgrade?
Hi Sterling,
Hmm... I would put it like this
"To upgrade to 502 successfully, install 502, unzip the patch, run repotools-<app> -clean, run the upgrade"
You are correct that the application is down during the upgrade. The upgrade will run repotools commands. The iFix in this case contains a fix for one of the repotools command. The repotools application "Installs" the patch (that is, loads the patched bits) in the same way the application will when it runs.
Does that address your concerns?
Ruby
Martha (Ruby) Andrews
2 other answers
It is better do one step per time. Finish the upgrade first, test it and then, apply the ifix patch.
This was for an upgrade from 4.0.7 to 5.0.2; rather than bring our multitude of applications online, install the patch, restart. In fact, I later got ifix3 and placed it in the patch file, excecuted our upgrades and started the application server. Seemed to work quite well.