CURL call for a triggered task dying after 15 minutes

[wani-amol]

All,

We are trying to run a triggered task through a CURL call and the pipeline runs for more than 15 minutes. The CURL call dies after 15 minutes.
We found that the Cloud URL has a 15 minute connection limit whether a connection is idle or not. Is there a way around this or to maximise the connection limit?
We have a few options including batching up data in order to run the pipeline in less that 15 minutes but was wondering if we could maximise the connection limit somehow?
Any help is really appreciated. Thanks!

-Amol

[kmiesse]

I’m able to get executions longer that 15 min to run with the On-prem URL but I end up getting 504 Gateway Time-outs not even 5 or 10 minutes into execution when triggering the task with either the Execute button in Manager or with a curl. The pipeline finishes successfully. Do you have any insight into that, @rsramkoski?

[kmiesse]

SOLVED: It was our load balancer having an idle timeout of only 5 minutes that caused the 504. We extended it and did not receive a 504 when trying the On-premises Secure URL.

[kmiesse]

@dmiller, do you think the Timeout doc https://docs-snaplogic.atlassian.net/wiki/x/4-MV should specify this 15 minute limit is only on Cloud URLs?

[dmiller]

I’ve forwarded this on to our Doc team.