The Zurich release has arrived! Interested in new features and functionalities? Click here for more

Integration fails intermittently in customer instance

Raj Rudra Chat1
Tera Contributor

Hi All,

Good day !!

 

I'm facing in our customer SNOW instance with one of the incident-to-incident integrations with a 3rd party SNOW instance. First, we raise a ticket through portal as per process, and then the incident gets assigned to the assignment groups automatically as per a custom business rule. Here, the expectation is that since the incident gets assigned to the assignment group which acts as a trigger, the integration gets triggered automatically through another custom business rule.

 

However, intermittently we see that in PROD, that doesn't happen. But adding a worknote helps to "nudge" the integration in those cases. We suspect it's a timing issue. We have seen in the logs that somehow the 2nd business rule doesn't pick up the assignment group the first time in those cases, even though we see it being set correctly in the incident.

 

I know I'm not providing any screenshots/code but please bear with me. Has any of you seen this issue ? If yes, did you put timeout() function anywhere ? This integration is a SOAP integration and I haven't seen much doc on SOAP anywhere. Please also note that this had been working fine for around a year since we migrated last year to a new endpoint on 3rd party's end, and just started happening around 2 months back.

 

Regards,

Raj

1 REPLY 1

TejasLogicX_11
Kilo Contributor

Hi @Raj Rudra Chat1 

Check order of business rules – make sure the rule that sets assignment group runs before the integration trigger rule. Use Order field properly (lower number runs first).  user order field to define the execution order of BR.

 

Use current.update() carefully – if assignment group is set and then another update happens, the integration rule might miss it. Try current.setWorkflow(false); current.update(); if duplicate updates happening.

Add condition check – in integration rule, check current.assignment_group.changes() instead of only current.assignment_group != ''.

Convert to async – if possible, use event or script action instead of direct business rule trigger. This avoids timing issues.

Debug in logs – add gs.log with current.assignment_group and gs.action.getGlideURI() to confirm when values are available