Open olblak opened 2 years ago
Can you share the manifest, so we can work on reproducing (and ensure it is not only a network issue) ?
I couldn't reproduce the issue reliably.
After a discussion with @dduportal today, we identify the root cause of the issue. It comes from the way we identify what should be clone in parallel. In the context of GitHub scm, we use the combo owner/repository if a repository need to be cloned, but when two similar scm configurations needs to be clone, we end up in a situation where different Go routines try to clone the same repository in the same location
https://github.com/updatecli/updatecli/blob/main/pkg/core/engine/main.go#L91
Ideally each scm should provide a function such as CacheID
so we could clearly identify if two scm configuration are similar enough to be clone once in the same location
Is there an existing issue for this?
Current Behavior
While testing a manifest with three git repository. I notice that from time to time that one git clone is failing with error message "context deadline exceeded"
Expected Behavior
It shouldn't
Steps To Reproduce
No response
Environment
Anything else?
No response