I am working on setting up the OPSLOG archive process for 12.2 and had a couple of questions. I currently have the archival process running in 12.1. I am in the process of upgrading to 12.2 and would like to take advantage of the new ARCHIVE set up.
1.The Admin Guide states: "Avoid reusing archives when you set the ARCHIVEHLQ parameter. We recommend that you rename any existing archives or that you change the ARCHIVEHLQ parameter before you start CA OPS/MVS." Does this mean that all of my previous archives will be overwritten if I use the same HLQ as the 12.1 archives?
2. Why can't I find the Archive Parameters in the Parameter Reference?
3. Based on question 2, is there actually an ARCHIVETRIGGER and ARCHIVETRIG parm in 12.2? I found ARCHIVETRIGGER in the parameter reference but not ARCHIVETRIG.
The new OPSLOG archive process suffered modifications after its first implementation. Please, review the current documentation at the CA Technologies Documentation site: CA Technologies Documentation
Here is a direct link to the updated instructions to implement the new archive process:
Archiving and Merging OPSLOG Data - CA OPS/MVS® Event Management and Automation - 12.2 - CA Technologies Documentation
And here is the link to the new Parameter Reference section regarding this subject:
OPSLOG Parameters - CA OPS/MVS® Event Management and Automation - 12.2 - CA Technologies Documentation
As you can see in the above links, only the ARCHIVETRIGGER parameter remains effective.
Before implementing this new process, please, make sure you apply the PTFs RO81809 and RO82494 for OPS 12.2. Visit the article from the link below where those fixes are recommended:
Also apply the PTF RO81217 if it is not applied yet.
Regarding the ARCHIVEHLQ, see that there is a note in the new documentation:
Note: Any pre-existing OPSLOG archive GDG base and Model datasets can be used in place of creating new ones. Update the ARCHIVEHLQ parameter with the name of the GDG base and the ARCHIVEMODEL parameter with the name of the GDG Model dataset.
Ok here is a follow up question. If I set the ARCHIVETRIGGER parm to 00:00 it sets the archive to occur at midnight each night. What about archiving a log that is getting close to its message limit? part of the reason I implemented the ARCHIVE process was to keep OPSLOGS from filling up. 90% of the time, midnight is a suitable but every now and then we get a ton of messages, like when someone turns on a trace, and the log fills up much quicker. How does the archive process handle that now if I want to do both a time and message count?
EDIT: After looking over my previous process, it appears that we could set it up this way:
Set ARCHIVETRIGGER to some number of messages. Allow the ARCHIVE process to do its thing when the log reaches said number.
Also perform a SWAP from one OPSLOG to another. This triggers message OPS4*3O which I have a rule that catches this event and runs the archive REXX.
While this process still currently works, will it continue to work in the future? I saw the time added to the ARCHIVE Trigger parm and hoping that the new process handled both a log filling up and a point in time archive.