new runtime check fails for docker build
stevie- opened this issue · 4 comments
Description
With the latest version 4.1.2 and 4.1.3 a new shell check for runtime
parameter was introduced. This breaks our workflows on github runner ubuntu-20.04 (or any runner which has not the selected runtime version installed). We still set python3.7
runtime which is not available on the runner, which is known and thus we use docker to build our lambdas.
It seems that this check is executed before the build process switches to the docker runtime.
Please check why this code is not working with the docker build setting.
- ✋ I have searched the open/closed issues and my issue is not listed.
Versions
-
Module version [Required]: 4.1.3
-
Terraform version:
v1.2.9 -
Provider version(s):
hashicorp/local v2.2.3
hashicorp/aws v4.36.0
hashicorp/null v3.1.1
hashicorp/external v2.2.2
Reproduction Code [Required]
{
create = true
create_role = true
function_name = "python3.7-docker-build-fails"
handler = "app.lambda_handler"
runtime = "python3.7"
source_path = [
{
path = "../lambdas/trigger-lambda/src",
pip_requirements = "../lambdas/trigger-lambda/requirements.txt",
}
]
build_in_docker = true
docker_image = "lambci/lambda:build-python3.7"
docker_build_root = "../lambdas/trigger-lambda/docker"
}
# requirements.txt
jsonpath-ng==1.5.2
Steps to reproduce the behavior:
< Are you using workspaces? > no workspaces
< Have you cleared the local cache (see Notice section above)? > clean cache
< List steps in order that led up to the issue you encountered > update module to 4.1.3 and rerun unchanged code
Expected behavior
The check for runtime binary should be done inside docker container.
Actual behavior
│ Error: External Program Execution Failed
│
│ with module.lambda.data.external.archive_prepare[0],
│ on .terraform/modules/lambda/package.tf line 10, in data "external" "archive_prepare":
│ 10: program = [local.python, "${path.module}/package.py", "prepare"]
│
│ The data source received an unexpected error while attempting to execute
│ the program.
│
│ Program: /path/to/bin/python3
│ Error Message: Traceback (most recent call last):
│ File "REDACTED/aws/lambda_function/.terraform/modules/lambda/package.py", line 1403, in <module>
│ main()
│ File "REDACTED/aws/lambda_function/.terraform/modules/lambda/package.py", line 1399, in main
│ exit(args.command(args))
│ File "REDACTED/aws/lambda_function/.terraform/modules/lambda/package.py", line 1198, in prepare_command
│ build_plan = bpm.plan(source_path, query)
│ File "REDACTED/aws/lambda_function/.terraform/modules/lambda/package.py", line 768, in plan
│ pip_requirements_step(pip_requirements, prefix,
│ File "REDACTED/aws/lambda_function/.terraform/modules/lambda/package.py", line 661, in pip_requirements_step
│ raise RuntimeError(
│ RuntimeError: Python interpreter version equal to defined lambda runtime
│ (python3.7) should be available in system PATH
│
│ State: exit status
Terminal Output Screenshot(s)
Additional context
I had the same issue when merging the latest version of code into my branch for this PR - #360
A fix is included in that PR (because I was unable to proceed without it) but it might be necessary to apply the same fix to a separate branch if I can't get the other changes included in that PR merged
Related question: What is the appetite for running docker during the plan phase? Obviously you would need docker, and it would need to pull the image...
This issue has been resolved in version 4.1.4 🎉
I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues. If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.