hey folks. i'm trying to get atlantis running for IBM in RedHat Openshift and i'm running into the following error upon container startup:

[dumb-init] /bin/sh: No such file or directory

i've tried just about everything I can think of, so figured i'd reach out to the community to see if anyone has any ideas what could be wrong. at the moment, i'm running the container as user 100 which should correspond to the atlantis user.

hey folks, just wondering if anyone ran into a problem where conftest would bomb out on plans with multiple projects? been running into an issue where conftest would fail to run. I believe this might be a race condition with

--update

flag since they all point to 1 directories for policies. Not sure if there's a way around this?

Hi folks, I'm having an issue setting up Atlantis. I'm using the GCE Terraform module to run Atlantis. I've configured the yaml, repos and webhooks. Atlantis starts up and exposes the web server. I create a pull request to a branch in the repo. I see in GitHub the Webhook was sent AND Atlantis replied with 200 and the body "Processing..." I have docker attach to the container and NOTHING is output when the pull request happens nor when the webhook is triggered (how did Atlantis reply processing??) Atlantis does nothing to the pull request. No comments with the plans or anything If I close the Pull request I do get something in the log: {"level":"info","ts":"2024-10-03T174135.939Z","caller":"events/events_controller.go:579","msg":"deleted locks and workspace for repo xxxxxx pull 5","json":{"gh-request-id":"X-Github-Delivery=bbe16c20-81ae-11ef-987e-56766d2edd8b"}} Anyone knows what can I check? It's not even logging when the pull request happens even tho it tells GitHub 200 and is in process, but it does log when I close the pull request

hey, does somebody found solution for https://github.com/runatlantis/atlantis/issues/2243 for terragrunt why it is important: • sometime we need to deploy apps which are consist with multiple modules and a few dependent module • plan will fail without moch_outputs (hard to understand what is really created), but fake values will apply with mock_output. Ideally will be that atlantis discard plan with previously generated fake values for dependent module and generate new one with real dependency module outputs.

Where do the

multienv

scripts need to live to work? I have it in my

atlantis

repo but does it need to be in the target repo?

running "plan-hook.sh" in "/atlantis-data/repos/company/platform-engineering/automation/companycm/650/default/terraform": exit status 127: running "plan-hook.sh" in "/atlantis-data/repos/delta-defense/platform-engineering/automation/companycm/650/default/terraform": 
sh: plan-hook.sh: not found

hi Have a question about the 'when_modified' field in the repo level config, we are using dynamic repo config generation and in the same way generating the var files for the project. The var files are not commited to the git repo, since it's not commited I beleive the when_modified causes an issue by not planning the project. Removing the when_modified field from config doesn't seem to help becuase of the default values. Do we have way to ignore this field and just plan the projects, regardless of modified file in the project

hi, sort of an odd question. but has anyone ever requested a feature to perform Atlantis on GitHub Orgs? Meaning we do exact thing as what Atlantis does now but consider an individual repo as a project? we are moving away from a monorepo approach and dividing things up among repos and it's becoming a pain to manage via Atlantis (administration of modules/version bumps, feature adds, etc.)

Hello I have a strange bug with atlantis. My repo is under a repo rulesets and when there is

atlantis/apply

as a required status check I am unable to run

atlantis apply

and I get the following error despite having all of the approving reviews from valid codeowners that were not the author.

Apply Failed: Pull request must be approved according to the project's approval rules before running apply.

Something that was interesting was when a repo is under just branch protection rules (different than rulesets but do similar things), it does not encounter this error. Also it appears that as long as the

atlantis/apply

is not a required status check it can run apply successfully. Would love to hear some insight that can allow

atlantis apply

to be ran while keeping the status check or any extra information that can help give me insight on why this error is happening? Thank you!

Does atlantis shallow-clone by default (when running with git/github)?

I noticed in 0.27.3 that i could use

atlantis plan -- -destroy

to destroy resources In 0.28.5 and later, it returns a regular plan instead of a destroy plan. Is this a known regression ? Edit: the issue was omitting the

$COMMENT_ARGS

in a custom

plan

step

Why does the following print empty string? How to reference

$REPO_REL_DIR

in the post workflow hook of

apply

?

post_workflow_hooks:
  - description: example
    commands: apply
    run: echo $REPO_REL_DIR

My Atlantis uses default

apply

workflow.

Does anyone using custom workflows to save terraform plans and applies to s3 for historical view ?

Does anyone know why a "environment" file is made inside the atlantis container under .terraform after an init? Its somehow messing with which statefile its picking from my blob storage remotely. The file only contains env:prod and then atlantis is somehow trying to use a state file per my backend config's "key" but suffixed with "env:prod" which doesnt exist. Heres my workflow:

version: 3
projects:
  - name: prod
    dir: terraform
    workspace: prod
    autoplan:
      enabled: true
      when_modified:
        - '**/*'
    apply_requirements:
      - mergeable
      - approved
    workflow: terraform_workflow

workflows:
  terraform_workflow:
    plan:
      steps:
        - run: rm -rf .terraform
        - init
        - plan:
            extra_args: ["-var-file=vars/${WORKSPACE}.tfvars", "-lock=false"]

Any guidance would be appreciated 🙂

Hi all, I facing two issues with atlantis on gitlab, I'd appreciate if someone can point me in the right direction: 1. I am looking for a way to disable repo locking while an MR is moved to Draft state, and again unlock if it's moved to ready without allowing the users to run atlantis unlock manually. a. Usecase is that people can plan their changes and move the MR to draft in order to unblock the pipeline and later allow them to continue working on the same. 2. I am wondering if anyone from the community is looking into this issue: https://github.com/runatlantis/atlantis/issues/4313. I am constantly facing this issue on gitlab where the pipeline gets stuck in processing stage and the only workaround is to close and recreate the MR. Thanks for your help.

Hey everyone, I want to help contributing this project. Could you help me out in linking some easy to get started with this project 👀

Did anyone else face autoplan / discovery issues in v0.28.x? Modifying

*.sh

scripts under stack's nested directory, e.g.

app/databricks-e2/databricks-workspace-analytics/scripts/global_init.sh

used to work and resulted in

app/databricks-e2/databricks-workspace-analytics

stack being automatically planned. With v0.28.x (I think) it stopped working and instead Atlantis attempts to plan:

app/databricks-e2/databricks-workspace-analytics/scripts

.

👋 Hi everyone!

Hello, welcome!

👋 Has anyone run into an error like the following? I don't think this is an atlantis issue, but after some searching thought y'all might have an idea:

* fork/exec /usr/local/bin/terraform1.7.5: argument list too long

I'm thinking we're running into a plan that is too large, admittedly the

ARG_MAX

size on the bottlerocket host is pretty small at

131072

Hello. When I add

prevent_self_approve: true

to disable the author from approving the policy I get the below error . Where should I add this flag ?

{"level":"info","ts":"2024-10-18T13:15:36.840Z","caller":"server/server.go:184","msg":"Policy Checks are enabled","json":{}}
Error: initializing server: parsing /etc/atlantis/repos.yaml file: yaml: unmarshal errors:
  line 20: field prevent_self_approve not found in type raw.PolicySet

Is there an atlantis comment that just reruns the policies ?

I cannot seem to get the

policy_checks

to work. I get this error. Has anyone else seen this?

1 error occurred:
	* policy_set: policies: conftest: Error: running test: load: loading policies: load: 1 error occurred during loading: stat cd: no such file or directory

Does anyone have any experience with terraform stacks and enabling parallel plan and apply without terraform workspaces? I have an atlantis.yaml like this

version: 3
automerge: true
delete_source_branch_on_merge: true
parallel_plan: true
parallel_apply: true
projects:
  - &template
    name: template
    dir: template
    terraform_version: v1.7.3
    autoplan:
      enabled: true
      when_modified:
        - "**/*.tf"
        - "**/*.tfvars"

  - <<: *template
    name: transit-gateway/us-east-1
    dir: terraform/transit-gateway
    workflow: us-east-1
  - <<: *template
    name: transit-gateway/eu-west-1
    dir: terraform/transit-gateway
    workflow: eu-west-1

workflows:
  eu-west-1:
    plan:
      steps:
      - init:
          extra_args: ["-backend-config", "env/eu-west-1/backend.tf", "-reconfigure"]
      - plan:
          extra_args: ["-var-file", "env/eu-west-1/terraform.tfvars"]
  us-east-1:
    plan:
      steps:
      - init:
          extra_args: ["-backend-config", "env/us-east-1/backend.tf", "-reconfigure"]
      - plan:
          extra_args: ["-var-file", "env/us-east-1/terraform.tfvars"]

However, this workflow does not allow parallel plan/apply to be enabled as Atlantis throws a state lock error

the default workspace at path terraform/transit-gateway is currently locked by another command that is running for this pull request.
Wait until the previous command is complete and try again

is this caused by the fact that both are using the default workspace?

We use local_file to write a file back into the git repo with some output values. Normally we just run this locally to generate the file so atlantis sees no changes, except in some cases the module running under atlantis seems to want to create the file, even though the file exists in the repo and on disk in the workspace atlantis is running under. Is there anything that would prevent atlantis from seeing files on disk?

I have an

approved_account_id

rego policy and I’d like to migrate the approved account IDs outside of the rego policy so they can be managed separately and be integrated into other policies. I’d like to grab the list of approved account ids from a yaml file in an s3 bucket and pass it into my rego policy. Is this possible in native

policy_checks

in atlantis ?

I’m testing out a PR with failed

policy_checks

and I cannot seem to get the defined list of users or team added to the failing PR.

<!here> We have started the move to the CNCF Slack so please do not post any more messages here and join our new slack community using this link https://communityinviter.com/apps/cloud-native/cncf

<!here> New post are restricted, please join the new slack to post your questions or help out.

the new channel in the CNCF slack is #atlantis

we archived

atlantis-community