Hi there,
We’re facing a random latency problem when using a Rest outbound connector task in multi-instances scenario.
This is a test case: we send 19 https rest calls to the same endpoint; each call has different payload. This same operation sometimes takes 4 seconds, sometime 1 second, sometime 20 seconds, which is weird.
Analyzing the worse case through Operate section, we see that usually 18 upon 19 calls returns within 1 second and the 19th takes 20 seconds. So we monitored arrival time at our endpoint of Camunda calls and they didn’t arrive simultaneously, but with a variable delay, from 1 to 19 seconds.
Also we prepared a test case in J-Meter: all 19 calls always arrive and are served within 1 second maximum, no latency, even when trying several times.
Is there anything we can inspect to understand why this is happening?
We’re using Camunda 8 SaaS, engine version Camunda 8.5+gen8.
In addition to my question: a different region from Camunda installation cluster (europe-west2) and our endpoint (europe-west8) might be the cause of this latency?
Hi @Anna.Cerbaro - it sounds like you may be on either the starter or enterprise SaaS plan? If so, I would recommend reaching out to support for this. I don’t think the region matters in this scenario; but the support team would be able to inspect your cluster and better assist.