Execution History: Expand execution history to keep more records

Related products: None

When you have rules that run multiple times a day, the rule Execution History (which appears to only keeps the last 50 executions) doesn't go back very far, which for rules that run hourly only show about 2 days worth of history. Please expand the Execution History so that you can view runs that span back farther.





@maksim_ovsyannikov Gainisght really needs improved logging capabilities. The ability to trace actions backward - regardless of success/fail - when end users report downstream issues days to weeks later are crucial to the admin role.





Rule Execution Logs from Scheduled runs not containing Action Details





Rules Engine: Extend expiration of Rule task processing files to 90 days





Salesforce Connector: Include detail of what happened in the logs even when successful



+1.

Yes at present the Rule Execution History is showing up the Latest 50 executions, is there any way to increase this limit.

When you have rules that run multiple times a day, the rule Execution History (which appears to only keeps the last 50 executions) doesn't go back very far, which for rules that run hourly only show about 2 days worth of history. Please expand the Execution History so that you can view runs that span back farther.

 

 


+1, this is a huge headache for us. 


@All,

 

How long back would you require to go back? We cannot store every execution, but we might be able to do better than 50. How about last 50 executions or last 7 days? 


Last 7 days would be a big improvement. 


We have a similar requirement. We would like to essential keep a history of last 30 day actions irrespective of how many times it runs. When we run rules multiple times a day - sometimes there are zero actions to take which we would like to filter out.

On rules that have no actions - we would like to examine the rule setup steps (occasionally for us) - like merge, transforms separately. 


In the lack of audit tables, and the need for users to have transactional level of accuracy of data, it is extremely important to be able to go back and tell the history of a change. Either that or develop ‘track fields’ in history objects like Salesforce does. 


30 days would be acceptable. This has been a huge need since I started working in Gainsight 2 years ago.


@rakesh any update on whether or not this is included in the upcoming Rules Engine horizon update?

Particularly for rules that run every 2 hours, this is a big issue because the rules history is gone within 3-4 days.  It makes an admins job very difficult to try and go back and figure out what may have happened when issues aren’t reported for several days.


I long thought that a downloadable CSV option that contains additional details and longer span would be incredible for troubleshooting/analysis.  +1 from me.


@rakesh as an example, I have a rule that runs every 2 hours.  As of right now, the logs only go back to Dec 17 at 4:17 AM:
 

I was on PTO last week, came back and was notified that this an account that should have been picked up by this rule on Dec 12th, didn’t.  But because the logs have already rolled off, I have no recourse to go back and investigate. 

Data issues don’t always reveal themselves due to a rule failure.  Data issues also don’t always reveal themselves instantly but perhaps days (or weeks!) later once a downstream dependency is identified as “off”


+1,000! Would love for this to be implemented. It would help save so much time troubleshooting rule partial successes and failures!


Hi @darkknight 

We are storing more data now in Horizon Rules. While our default behavior is same as 50, it is now more straightforward to expand this.

As we look deeper, there is one problem that remains in last x days mode: If there is a rule that is configured to run infrequently - once a week or once a month, that would have only 1 or no executions in execution history. Will this be a problem realistically in how you use Rules? 

 


@rakesh we definitely have several rules that run with less frequency, so would ideally need longer history for those time-wise.