This message was deleted.

yes, that happened to me because I was out of resources, memory or cpu

but it's consistently failing through atlantis service, it always passes if I go into the container and git clone, and go into the directory and run

terragrunt plan

myself If the service runs it, it gets an error that some attributes aren't found

Atlantis adds overhead to those commands

it has a good amount I thought:

resources:
          requests:
            memory: 1024Mi
            cpu: 1024m
          limits:
            memory: 2024Mi
            cpu: 1024m

that is very strange

it's not a big a repo, it's like 30 files I'll nuke this pvc and have it recreate though just to test

ok

Okay, node is back up and just saw it clone again here:

atlantis@atlantis-0:/atlantis$ ls
atlantis.db  bin  lost+found  plugin-cache
atlantis@atlantis-0:/atlantis$ ls 
atlantis.db  bin  lost+found  plugin-cache  repos

The issue is almost instant though, its commented already the plan is failed:

data.terraform_remote_state.project_base.outputs is object with no attributes

if I go into that atlantis directory,

/atlantis/repos/

and get into the

project_base

directory and run, it tries to create/set the outputs as if they're not already set however If I go into the

/tmp

directory and clone myself and got into

project_base

I see "No changes" and no errors will occur when I got into

project

and run terragrunt, I get an output with 6 resources to create

could it be a problem with access of the state?

but then how could I go into /tmp/ and clone/run successfully?

same commit hash?

yes and same container, i'm just exec'ing into it kubernetes pod

same branch etc?

yes same branch too

that makes no sense

good idea, will try that

and this happens to a particular PR?