-
Notifications
You must be signed in to change notification settings - Fork 17
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add install of .NET 3.5 #1585
Add install of .NET 3.5 #1585
Conversation
Adds install of .NET 3.5 for building the MSI installer
Thanks @slide ! Non blocking comment: do you think we could add a test in https://github.com/jenkins-infra/packer-images/blob/main/tests/goss-windows.yaml? |
@slide the build did fail during the dotnet installation. While we can discuss giving you access to the build logs, here are the errors:
|
Hmmm, that doesn't really give any useful information that I can see. Let me do some digging. |
Try different install method
Dump the log file from DISM
@slide , while @smerle33 is helping you on matrix with the logs you added, please be aware that he caught the following issue: jenkins-infra/helpdesk#4445 causing build issues |
For info: I've updated this PR with:
=> let's see the result |
# install .NET 3.5 for MSI build | ||
& "choco.exe" install dotnet3.5 --yes --no-progress --limit-output --fail-on-error-output | ||
if(Test-Path "C:\Windows\Logs\DISM\dism.log") { | ||
Get-Content "C:\Windows\Logs\DISM\dism.log" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Installation keeps failing in the context of chocolatey with the following DCIM error:
https://gist.github.com/dduportal/3a725124ab0d88184980ef72a9685892
It's weird because the Goss test responds True
. Is there already a 3.5 .NET install in the system images?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not sure, I guess if the goss test is getting true, then it is?!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That is the question :)
- I assume this test should fail on the current VMs on ci.jenkins.io (I'm double checking)
- If yes (it fails today in production) then it means that the chocolatey install in this PR does something to make the test pass. Even if the chocolatey install "fails" (kinda).
- We could stop caring and go with it.
- Or we could find the proper flag to pass to chocolatey to avoid the error at all (better but heavier in term of effort).
I'm currently focusing on the Windows in EC2 + fixing the goss test errors on Windows for hadolint
CLI. As part of the debugs (I am building an instance with the usual packer process but I can access it to debug it in remote desktop), I'll check this.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I confirm that, in production today (version xxx of the image) , the test answers False
. It confirms your chocolatey install does what we want (looks like).
Gotta check the DSIM log to check if the error isn't as minor "post install step" which we could ignore or disable through flag.
(I'm working with #1611: I'll try to see if I can check on interactive machines) |
Looks like this should be good. I'm merging and we'll start with a first release so you can check if NET 3.5 is ok. |
Adds install of .NET 3.5 for building the MSI installer