When i run a process from orchestrator. It randomly get stuck in between. what could be the possible reasons. because i never faced issues for this process while running from studio
Hi @rizvana.mohammed
Pls share what is the exact failure? or
Process may get stuck in Orchestrator due to unattended session not unlocking, missing app permissions, different screen resolution, selector failures in unattended mode, or application load delays on the robot machine.
Happy Automation
It is getting stuck on Azure Open API call. It is not like in first call run is getting stuck but after 10 or 15 calls
If it hangs after 10–15 Azure OpenAI calls, it’s most likely hitting rate limits or throttling. The request doesn’t fail, it just waits. Make sure the API call has a timeout, add a small delay between calls, and use retry logic.
How are you using azure here http request or Azure open ai activity…
Can you try adding log messages before and after activity azure..
but it works good from studio. Also i don’t see any timeout parameter for Microsoft Azure OpenAI HTTP Request activity. I have given retry also.
I am using Microsft Azure OpenAI HTTP request. i have provided the logs before and after the activity and came to know it is getting stuck on this particular activity. the main issue is it is not throwing error also.
Got it.
You can set an explicit Timeout (e.g., 60000–90000 ms) on the HTTP Request activity.
Then wrap the call in a Retry Scope with limited retries.
Also add a small Delay (1–2 seconds) between consecutive calls.
Try to avoid parallel requests and reduce token usage per call.
I wrapped in retry but it is throwing error for few times but after that process is getting stuck. could it be orchestrator connection issue
From the admin page, can you check if Integration Service is enabled for the tenant where you are running the process.
If a process runs fine in Studio but randomly hangs when triggered from Orchestrator, common causes :-
- Unattended execution runs in a locked session, so UI activities depending on screen focus or resolution may behave differently.
- Orchestrator executes faster than Studio, which can expose missing waits, element checks, or retry logic.
- Machine-level issues such as high CPU/memory usage, Windows updates, or unstable sessions during execution.
- Differences in assets, credentials, or environment variables resolved at runtime.
- Insufficient logging, making the job appear stuck instead of failing - review Orchestrator logs to find the last successful step.
Yesthe integration service is enabled
