When doing solution imports using the Build tools tasks to first Import a solution, and in the next step Upgrade that solution, the Build tools tasks report an error, and if you are hitting refresh in make.powerapps.com quickly enough, you'll also see the error. But then like magic, make.powerapps.com reports success. That leaves me with a pipeline that fails however due to the error being reported back to the build tools.
The error is:
This happens quite frequently lately. Like I said, I'll see the error reported temporarily in make.powerapps.com when looking at (and refreshing) the solution history, but it then changes within a few seconds to "Success".
Anybody else experiencing this?
I am currently experiencing this. Can I ask how long your Apply Upgrade (PowerPlatformApplySolutionUpgrade) step takes before timing out?
Can I also confirm that you are running this task aysnc? What timeout have you set?
Did you have any luck resolving this?
I am running async.
My timeout is set to 3 hours.
My pipeline task is appearing to fail around the 90 minute mark but the failure of the pipeline task is the same duration as the uninstall job in make.powerapps.com that is successful.
No luck in resolving this. I do have the task running async. The time out I have set to 60 minutes but I don't think it has anything to do with the time out of the task in DevOps. This error is reported by the API doing the actual upgrade. (You can see this error displayed for a very short period of time if you are refreshing make.powerapps enough.) I think the solutions are: the team responsible for the power platform build tools needs to trap that particular error and not let it bubble up or work with the team responsible for the actual upgrade actions for the dynamics platform and have them investigate why that error is being surfaced, then caught and handled properly by make.powerapps.com.
I agree Jeff
I have a support call open with MS now and waiting on the product team to get in touch. Did you raise a ticket yourself? Did you find a nice way to handle the failed step in the pipeline? I have the feeling I will have to set condition: always() for the tasks that follow the "Apply Upgrade" task until MS fix the build tools. I might see if the Wael Hamze tools handle the failure more gracefully.
I didn't raise a ticket and have not found a graceful way to handle the error. It's one of those things I've just been dealing with. Let me know what you find out from your support ticket. Thank you.
+1. We first started encountering this last week; have never encountered it in thousands of deployments prior to then, but it's happening consistently now.
Our 'Apply Solution Upgrade' build task is set to apply the upgrade asynchronously, and the maximum wait time is set to 60 minutes; the failure always occurs prior to that time being reached.
@Rob_Nightingale - as Jeff said, would be great if you could let us know how you go with support.
We are also seeing this error now too. It first started in a single environment (our QA environment) but it's now happening somewhat randomly in almost all environments now. I have also opened a ticket for this and referenced this blog post in the MSFT ticket.
Use the PP Build Tools they said. It will be fun they said.