Panfactum / stack

The Panfactum Stack
https://panfactum.com
Other
16 stars 5 forks source link

fix(#95): use integers for cpu_milicores above 999 #96

Closed wesbragagt closed 3 months ago

wesbragagt commented 3 months ago

This PR sets the resources cpu divided by 1000 as when setting it to 1000mi fails from the provider level.

After applying in my cluster I see this in the pod definition image

https://github.com/Panfactum/stack/issues/95

fullykubed commented 3 months ago

I believe the issue at hand is the serialization format used by the kubernetes_manifest resource. This manifest is that last one that we haven't converted to a kubectl_manifest b/c it involves some manual migration steps. Given the sensitivity of this resource, I wanted to take the time to thoroughly test and document the migration path.

wesbragagt commented 3 months ago

I believe the issue at hand is the serialization format used by the kubernetes_manifest resource. This manifest is that last one that we haven't converted to a kubectl_manifest b/c it involves some manual migration steps. Given the sensitivity of this resource, I wanted to take the time to thoroughly test and document the migration path.

Totally understand @fullykubed. Thanks for the info. I'm unblocked as I've cloned the module and edited myself.

wesbragagt commented 3 months ago

Closing since you said that it's solved by using the new manifest definition.