It depends where in the job it fails.
For example, what logs are involved with an SLP backup and duplication - answer about 25 (approx). But this includes, backup (nothing to do with SLP really), the SLP 'management of the image', and then if that all works, the duplication (again this part has nothing to do with the SLP side - well a tiny bit as nbstserv has to generate the correct command ).
How many logs do you generally need for a true SLP problem - usually 1, nbstserv log.
Advice, take each issue as it comes, and work through it depending on where in the job it fails.
Activity Monitor usually shows the error and the process that it failed on. For example, a backup might fail with an error in bptm. The error might be in bptm, or the actual error might have happened in a process related to bptm and just gets 'reported' in bptm.
Then use the process flow to find out what the main processes are that bptm talks to - again this depends where in the job it fails, but as a high level overview:
nbpem starts the job, passes details to Activity Monitor, nbjm then talks to bpjobd and only then does the job appear in activity monitor. nbjm then starts bpbrm which in turn starts bptm and bpbkar.
bptm asks for resources from nbjm, nbjm talks t nbrb and nbrb talks to emm and mds to get these.
If all that works, bptm then talks to ltid to get he tape loaded, ltid goes off and talks to tld which talks to tldcd to get this done ... and so on ...
So you can see, you ask an impossible question - add into the above communication loogs, and OS stuff and you have a load of logs and it would be very rare that we needed all of them.
Better is to learn 'how' to collect and setup logs (many people cannot do this), and how to work out the approximate area of the problem. Fair to say, sometimes it is not obvious and you have to hunt around, but for 'common' issues you can usually get quite close.
