Performance with Triggered task

As part of our project requirement, we are using Triggered task and observed that performance is not as expected.

So, can someone help to understand whether the trigger task has some latency in general to complete the transaction?

What is the expectation, and what are you observing?
Where is the pipeline being triggered (Cloud URL or Ground URL)?
Where is the pipeline being run (on a the cloud or Groundplex?)
How long does the pipeline take to run if executed manually?


In our pipelines we are using less snaps which expected to complete in less than 10 sec but we observed it taking more than that. We also observed it taking expected time whenever we run it manually. We are using Cloud URL and some times Load balancer URL for triggering. The Pipeline running on Coudplex.

I’m facing the same issue triggering the task by Cloud URL. A Ground-Plex is used to execute the pipeline.

Task triggered using the Cloud-URL: 3,5-8 s
time_namelookup: 0.001784s
time_connect: 0.163308s
time_appconnect: 0.496279s
time_pretransfer: 0.496336s
time_redirect: 0.000000s
time_starttransfer: 3.937991s
time_total: 3.938045s

Task triggered using the OnPrem-URL:
time_namelookup: 0.000699s
time_connect: 0.003258s
time_appconnect: 0.028382s
time_pretransfer: 0.028439s
time_redirect: 0.000000s
time_starttransfer: 0.237001s
time_total: 0.237062s

Triggering a task through a cloud url and ground url are now substantially different. The flow of a cloud triggered task will go through the SnapLogic Control Plane, where as the On premise URL has been optimized to try and avoid making any requests to the Control Plane when possible. Another difference is that any data passed in or out will be proxied through the control plane with a cloud url. Where as with the on-premise URL, your HTTP client connects directly with the Plex. So the on premise URL should have a substantially lower latency in most cases, especially if executing on a ground plex that’s co-located with the client.

The Cloud URL is convenient to use when the client might not be on the same network as the Plex as it allows connectivity, but if you can use the on premise URL, then that’s generally a better way to go.