Online Ultimate Insights


ARCHIVE Facility

Question #1w

Archiving without CULLing

We're thinking about using the Archive facility of NetResults/z to archive our Netlogs. We've been experimenting and have noticed that some messages are "not" written to the archive output GDGs.

We're aware of the CULLing process and how it distills the data, but we would like to control it.

How can we ensure that the data we need, will be on the output archive data set ?

Answer #1w

Before answering the question, it's probably a good idea to first "define" the term CULLing. We use the term CULLing to mean "removing that which is worthless", the term CULLing has also been used, by others, to mean just the opposite.

Ok, now that we understand the term CULLing, let's repeat the above question;

How can we ensure that the data we need, will be on the output archive data sets ?

The answer to that question is twofold; You can either "suppress" Culling, or you can "tweak" the NetResults/z Message Profile Table to give you "exactly" what you need.

To suppress Culling, is easy; Just add the "NOCULL" operand to your Archive request, and "everything" that has not yet been archived, will be written to the output archive data set.

Tweaking the NetResults/z Message Profile Table (MPT), however, is a bit more involved. As released, the MPT is "already" fine-tuned. Using it, as is, to CULL your Netlog archive data sets, can yield an output data set reduction of 40-80%; quite substantial.

The release version of the MPT, by default, CULLs the following:

All, but the "original" copy, of ASSIGNed messages.

All "C", "U" and "*" type messages.

Many "non-informational" type messages; of which there are many.

Of course, any of the above CULLing rules can be easily negated. Simply change or remove the rule from the MPT.

So, what should you do ?

Well; if you've got voluminous Netlogs (like those shops that replicate Syslog messages on Netlog, or those shops that do an outrageous amount of ASSIGNs); then tweak the MPT. The benefits, DASD-wise, can be substantial.

Subsequent processing of Culled Netlogs is also much faster.

If, on the other hand; you've got small Netlogs (an oxymoron ?) or you want to archive, everything, regardless; then simply suppress Culling by specifying the NOCULL operand.


[Home] [About Us] [Products] [News] [FAQ] [SYSLOG Viewing] [NETLOG Viewing] [LIST Facility] [PROFILE Facility] [SHOW Facility] [ARCHIVE Facility]