https://cloudposse.com logo
Join Slack
Powered by
Hello, When running `atmos terraform deploy` or `a...
# atmos
b
Hello, When running 
atmos terraform deploy
or 
atmos terraform apply --all
with concurrency > 1, we see atmos ignore the provider/module cache in github and pull the provider for every component. Is that expected behavior?
e
Not expected; sounds like a bug we had fixed. Or maybe it hasn't merged yet. Let me check
I assume you're on the latest version?
b
yes, 1.225 I believe
e
also, for additional context, are you running this in CI or locally
b
github runner, we actually crashed our runner when it tried to pull the provider for ~90 JIT vendored components
e
lol
Also, reminds me, wish we published our components as OCI artifacts
Ok, 2 things.
The terraform native provider cache is not concurrency safe; so atmos automatically disables it.
Separately, there's the atmos native caching for providers. That is concurrency safe.
The atmos registry proxy should not be disabled by concurrency; are you using that one?
Copy code
components.terraform.cache.enabled: true
b
I thought we were using the atmos native cache, but let me check. I know my coworker got around it with: Restored that cache dir with this config in `atmos.yaml`:
Copy code
rc:
      enabled: true
      plugin_cache_dir: $HOME/.cache/atmos/terraform/plugins
      plugin_cache_may_break_dependency_lock_file: true
e
That will restore the terraform cache dir, the same one terraform says is not concurrency safe 😉
b
so I'd say the answer to your question is no, I'll see if we can test the atmos cache later
e
also, 
$HOME/.cache/atmos/terraform/plugins
is home interpreted by Terraform? (I don't think it's interpreted by atmos)
i mean the 
$HOME
b
I think it is by tofu
2 Views
Open in Slack
PreviousNext
Powered by