Hi Duane,
Assuming the target system was built with an option 21 save then restore, all the apps and paths are there and ready to be started. But what to replicate can be a 'per application or service' dependent response.
For instance, reference this IBM document posted regarding the replication of the 'Content Manager'.
https://www.ibm.com/support/pages/system/files/inline-files/CMOD%20for%20i%20-%20Considerations%20for%20High%20Availability_0.pdfMost of the items in this path are user entries and can be easily replicated and used on the target system without any additional intervention. However we see from the document that some apps are concerned that replication may lock some of their objects during replication.
Some clients walk through the contents carefully to identify specifically what they want to make available on the target node in real time and select those in an IFS replication group. Entries in the HTTPA and HTTPSVR require more handholding to get new applications mirrored from the primary registered onto the target node(s) when they are changed or added on the primary system. (Just because the objects are replicated doesn't necessarily mean they will be available to start using the HTTPADMIN without making certain all the requisite objects are also replicated.) In other words, you can add a HTTPSVR from /qibm/UserData/ that is actually located in /www/some_child_folder.
Another option is to set up periodic refresh groups to snapshot the contents of this folder and dependents to keep contents up to date while avoiding locking objects during critical operations of the production app. But object locking can also be avoided using IFS journaling to replicate object changes while setting an appropriate object replication delay at the replication group level to avoid application conflicts.
Hope that helps.
------------------------------
Mark Watts
Software Engineer
Rocket Software Inc
Waltham MA United States
------------------------------
Original Message:
Sent: 10-13-2021 12:35
From: Duane Gingerich
Subject: Replicating /qibm/UserData
Just wondering how people handle the replication of /qibm/UserData?
------------------------------
Duane Gingerich
SE
Able-One Systems Inc
kitchener ON Canada
------------------------------