db_cluster_parameter_group_family issue

[fir1]

Hi,

I suggest the parameter from terraform files db_cluster_parameter_group_family= "aurora-postgresql11" must not be hard coded.

github-actions-deploy-docker-to-ec2/operations/deployment/terraform/postgres.tf

Line 65 in ffc6f48

db_cluster_parameter_group_family = "aurora-postgresql11"

Instead I kindly suggest if we can move this to the RDS Inputs of Github Actions (https://github.com/bitovi/github-actions-deploy-docker-to-ec2/tree/main#rds-inputs) and pass down to the terraform files as environment variable ?

The current implementation will force you to use only in postgresql11. Anything above >11 version will not work which breaks the point of having aws_postgres_engine_version RDS input for the Github actions logic.

Thanks,
Fir

[LeoDiazL]

Hi @fir1

Thanks for your feedback and getting involved!
Let me look into it and see what are our possibilities here. We really like to add support for every possibility, but sometimes it's really complicated to tie everything together.

Leo

[fir1]

Hi @LeoDiazL ,

Thanks for the great work, i do understand 👍 .

[fir1]

Hi @LeoDiazL ,

I have made PR , have tested it looks like it is working.

If you could check it out and if all good would be good to merge it.

#123

[fir1]

Hi @LeoDiazL,

I have observed a behavior in this particular GitHub Actions package. The described use case is as follows:

1. When files are updated or new files are added to the GitHub repository, it triggers GitHub Actions. As a result, the content in AWS EC2 is updated to reflect the changes from the GitHub project, and a new Docker image is created.

2. However, when files are deleted from the GitHub repository, it triggers GitHub Actions. Unfortunately, the corresponding files are not removed from the project's content in AWS EC2. Consequently, GitHub Actions generate a Docker image that is not up to date, and in some cases, the project fails to start entirely.

More precisely project which i am testing is written in Golang with relevant Dockerfiles, and docker-compose.

I would greatly appreciate your assistance in resolving this matter.

Thank you in advance.

[LeoDiazL]

Closed with PR#123
@fir1 will look into that Docker issue.
Ansible execution of docker-compose should be recreating the image from scratch each run.
Might add a docker-compose down --rmi all --remove-orphans somehow to ensure we are cleaning everything up.

Just in case, You can get help or ask questions on our Discord channel!