Skip on download failure

(imported topic written by sinucus)

Does anyone know if it is possible for Tivoli to skip an action after a timeout period? Specifically I’m looking at the baselines. Example;

I have 5 fixlets in a baseline. Fixlet 1 runs fine, Fixlet 2 needs to run patch.exe but it is not available for download, so it just sits there retrying every 10 minutes until the original action times out. Fixlets 3-5 never get run. Can I force a timeout for searching for files to download so that it will run fixlets 3-5 without having to start a new action?

Thanks

(imported comment written by CharlieD)

I also would like to know how to do this. I’ve had an issue when a manual cache item has been purged that choked up a baseline.

(imported comment written by sinucus)

Just to clarify, this is a situation where the Tivoli server is airgapped and unable to physically download the patch, it will never work. I would like the download action to just stop after it has tried and failed.

(imported comment written by jgstew)

I think you can lower the timeout between download failures and lower the number of retries for download failures. This would atleast cause it to fail faster instead of hanging.

(imported comment written by sinucus)

Do you know where this setting is? Or what it’s called?

(imported comment written by sinucus)

OK, so I found some settings
here
. The two that looked the most promising were these…

_BESGather_Download_RetryMinutes

_BESGather_Download_RetryLimit

Now the problem is they did exactly that. I changed the retry time to 1 minute and the Retry Limit to 1. So it waited 1 min, retried and failed and then… stopped… It’s just doing nothing, like it always does. I cannot find a setting to have to client/server skip a fixlet/subaction on failure…

“Run all member actions of action group regardless of errors” IS checked.

Thoughts/Ideas?

(imported comment written by jgstew)

I didn’t think this would fix the issue, just fail faster, which is slightly better.

I’m not sure it is possible to skip an action due to failed downloads, but it does seem like that should be what happens when you have the run all Member actions regardless of errors checked.

(imported comment written by vpetrell)

It isn’t possible to do based on all the testing and research we have done. We have this issue when patching SLES, because Novell will sometimes change their content but IEM content isn’t updated fast enough, so the downloads just keep retrying and the baseline gets stuck. Same technical issue as your air gap scenario.

We have an RFE to address this and also raised this with Product Management. Up vote RFE for support and maybe we will see a solution for this faster.

http://www.ibm.com/developerworks/rfe/execute?use_case=viewRfe&CR_ID=34736

(imported comment written by sinucus)

Thanks for the link. I have voted for your RFE. Let’s hope they do something about this.