Acknowledged

Auto Re-run Rules that Timeout

  • 15 September 2016
  • 7 replies
  • 50 views

Userlevel 5
Badge
Rules within Rules Engine periodically fail with the following error:
- Error: Your query request was running for too long.
- Error Code: QUERY_TIMEOUT

It would be great with Gainsight Rules had a re-try feature when failures like this happend where you could retry up to X times every X minutes

7 replies

Userlevel 7
Badge +2
We ran into issues like that too Jeffrey.  Salesforce has a hard timeout limit so if you are parsing a very large amount of data via a SFDC or Native GS object, you'll probably continue to get timeouts. Workaround for us was to try and segment the rules in such a way where we were parsing smaller amounts of data per rule.  Kind of a pain, and caused us to have to duplicate some rules, but it helped eliminate the timeouts for the most part. 
Userlevel 5
Badge
Thanks for the advice.  We had many of these timeouts when were first started using Rules Engine but over time we were able to make adjustments to the filtering as you suggested.  We rarely see this issue anymore but it does occur occasionally even on queries that typically run extremely fast. We have other ETL type tools that we leverage that provide this retry type functionality and it would be great if Gainsight could consider adding it as well.
Userlevel 6
Whenever a SFDC call fails with timeout, Gainsight does a retry of the same call after 10 seconds before it reports the timeout failure. 
It could be a case where SFDC servers were under load / down. 
Userlevel 5
Badge
We had a timeout on a rule today that has never timed out before. 10 seconds is probably not enough to wait.  It would be great if you could put this control in the hands of the configuration so that the retry for more important rules could be delayed longer than other rules.
Userlevel 7
Badge +1
This reply was created from a merged topic originally titled Attempt to rerun rule on error.

Have a situation where a rule failed, but not for any given reason.  It was a complete failure, though the rule had run perfectly fine the day before and has been running fine since.  Would like to see that, when the error isn't specific to data problems (missing Account ID's, etc), that the rules engine will try to rerun the rule at some specified time later, maybe after 60 minutes or so?  Especially if this rule is part of a rule chain, or even just part of a logical ordered rule run, the failure of this rule can cause other failures.
Userlevel 7
Badge
Hi There,

I recently received a request from one of the user regarding this and I could see that this has been marked as implemented. May I know why this has been marked Implemented?

Userlevel 7
Badge +2
@shiv_kumar_katiyar Thanks for letting us know, changed the status.

Reply