Discovery experts, which error table do you use? (Discovery Active Errors vs. Automation Error Messages)

[Jay Harris]

Hello, I am creating an process for my organization to better manage discovery errors.  During my discovery I have been reviewing the error reports from the CI Schedule Manager and the Discovery Dashboard.  Discovery Dashboard uses the discovery active errors table and the CI Schedule Manager uses the automation error messages table.  They are both useful, however I don't understand why there are two tables.  Can someone explain the difference between the two and use cases where one is more useful than the other? 

[DaveHertel · Accepted solution]

Hi -- I never use the disco dashboard.  While its pretty and simple, it gets in the way of efficient troubleshooting IMHO.  For the reason you point out (2 tables... similar errors) its doesn't help beyond providing big bold numbers... which are misleading in my opinion.  Further, drilling into the error tiles and details behind the dashboard leads teh user down the wrong path (again, IMHO).. because it doesnt dive into the depths of the logs you REALLY need to troubleshoot.     I always recommend avoiding it - its visual 'appeal' is much less appealing when you get frustrated trying to troubleshoot whats behind the numbers.  Further, I find the summary numbers misleading and conflicting.  What is called an "Error" on the dashboard isn't always an "error" in the log or vice versa.   That is poor design by degrading and confusing the troubleshooting efforts conflicting terminology.  

For disco admins that are focused on the elimination of errors (not just viewing a pretty summary), I leverage the REAL Disco tables where the 'action' is. They are more effective but not always clean/simple in interpret.  ECC_QUEUE, Discovery Device history & Discovery Log are my go-to troubleshooting sources.  While the *automation tables used by the dashboard are a waste of time for me.  A few reports wrapped around useful disco tables are much more efficient for digging into errors and ultimately reducing them.  My 2 cents

Hope this helps?

[Ashutosh Munot1]

HI,

Agree. 

I always use discovery logs, patterns logs and ecc_queue for debugging.

I only use this two tables when i need a dump in excel which i need to share with teams.

Thanks,
Ashutosh

[Jay Harris]

Thank you both.  Your insight is appreciated.

[Jeff Mayrand1]

Have either of you tried using the automation_error_msg table?  We are looking at this as well and are finding a lot of the relevant information there.  The dashboard certainly could be a lot better.  Having the hostname of the attempted discovery for instance would be a great start.   Also the consistency of the error messages etc could be a lot better IMO.