[Question]: Crashing Compile-AppInBcContainer/Publish-NavContainerApp in a PR pipeline - how to keep the container alive after crash for further investigation? #1595
Replies: 13 comments 2 replies
-
Beta Was this translation helpful? Give feedback.
-
Sorry, didn't see this - could you include the full log? |
Beta Was this translation helpful? Give feedback.
-
all good - what confuses me is this $tempZip is a path to a file (alc.zip), the Copy-Item uses this as Destination?! ah now! |
Beta Was this translation helpful? Give feedback.
-
when I add some -verbose to this currently I see: than it wants to extract: VERBOSE: Performing the operation "Create Directory" on target "Destination: C:\build\vsix". is this some winrm, process isolation thing, pwsh7 thing? |
Beta Was this translation helpful? Give feedback.
-
this looks like it is getting somewhere: I added $bcContainerHelperConfig.useWinRmSession = "never" to my config, and now the compiler folder was made and also the compilation process with alc.exe started to execute! what would be the best spot to place this setup in AL:GO if it is really fixing my issues? "fingers crossed" |
Beta Was this translation helpful? Give feedback.
-
that helped! without winrm the process runs. So looks like host to container have issues to keep the connection alive... |
Beta Was this translation helpful? Give feedback.
-
so @freddydk - this winrm is the thing that makes it go boom. All runs I triggered now without it worked flawless. About the build agent: it is based on the Create a self hosted runner runner document. Right now the basic D4x, choco, and the installation steps mentioned in the howto. To verify this once more: what is the most appropriate place for the setup to tell the whole thing: make not use of winrm? |
Beta Was this translation helpful? Give feedback.
-
You can add "useWinRmSession": "never" to .github/AL-Go-Settings.json but... - I think the problem might be something else - if you could include the full log, I might be able to spot the problem. Also, which user is running the agent on the runner? |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
System is fine - please email a link to where i can download the log - I cannot say a specific thing - there is a ton of info in the log |
Beta Was this translation helpful? Give feedback.
-
Did anything ever come out of this? We're currently seeing the same error in multiple repositories. The WinRM setting does seem to fix it, but I can't imagine that's the actual fix. |
Beta Was this translation helpful? Give feedback.
-
The real fix is that BcContainerHelper itself figures out that it shouldn't use WinRM - will add that to the list of things to get done |
Beta Was this translation helpful? Give feedback.
-
Looking at this. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Question
Hi,
I am about to add a test app to one of our PTE apps. In order to make this work I needed to get rid of "use CompilerFolder:true" and "doNotPublish: true".
So I can see now, the System is spinning a container, once it wants to compile it just crashes the alc.exe :
This internal error is not visible anywhere to me, I do get - thanks for that, the container logs, but those are looking quite strange "out of the box".
(I have to flip to the Details tab and XML view and read the EventData there as the "General" tab shows quite often nonsense)
Strange as I am, I connected to the Build Agents machine, started a container with the same Image, and called Compile-AppInBcContainer manually from the prompt...and "surprise" that works just fine ...
IF I now reenable CompileInFolder, I actually pass the compiling in container issue , however, once it wants to install the armada of tests apps from MS, it just crashes there on publishing apps ( feels a bit like "out of memory" - here the Events in the container log shows Warnings , EventID 705 and 712 ). However, If I do this manually from the powershell, against my selfmade container it just works :/
The Agent was build with the current ARM template. I had a lot of noise and memory consumption from the "Antimalware Service Executable" (real time protection (turned it off)). I also resized the VM to something more memory friendly (B8ms 32gig). that helped me to get "a bit further" but still crashing while publishing test apps:
So, I am a bit puzzled why the pipeline container just crashes the alc.exe (and can't handle the import) while basically the same flow from powershell works. Could this be a permission/security thing? I mean, powershell runs with the vmadmin, while pipelines are executed by a windows service. (still, why would that hit the alc.exe within a container?)
Can I keep the container that the pipeline creates? I recall a parameter ... maybe?
Also some other hint to get this working would be fine :) (I think I will try to make use of another host VM...just to see what will happen.)
I know, this is not made for on Prem, but still the import of MS TestApps should just work, it has not even started to publish my apps. Something is fishy with this container ....
some more details: Al.Go: v6.0
PTE Project "for OnPrem" BC v23
container isolation process
as long as no container/testApp is involved, I can compile and deploy to hosted environments.
Agent runs in a VM on Azure made with the ARM template
can Repro? Yes - can I share? no - at least not public :/
Beta Was this translation helpful? Give feedback.
All reactions