#terraform-aws-modules - April 2020

Hello, requesting some thoughts/opinions. So I am looking at building reusable tf modules for my teams. We are divided in our opinion on whether we should use a mono-repo and have sub-directories for each aws resource or one github repo per resource. On my previous projects, I have done the later. Each component has it's own lifecycle (tagging etc. ) that way and also only the required modules get downloaded and not the entire mono-repo during terraform get. What do you guys think ?

i agree with separate repos from an access perspective, especially with different teams. You might not want some teams accessing/building vpc or iam resources, while letting them build out ec2 or s3 etc...

Hello - I am using the terraform-aws-alb module and am trying to figure out how to attach targets to the created load balancer. I have instances that are running due to the autoscale_groupmodule, but I’m uncertain how to attach them. I’ve looked at the regular Terraform aws_lb_target_group_attachment resource, but haven’t worked out how to deal with the fact that I have two instances but target_id on aws_lb_target_group_attachment appears to only take one id. Any guidance would be much appreciated.

Are you using the https://github.com/terraform-aws-modules/terraform-aws-autoscaling module? I don't see any module named autoscale_group on the main module registry.

If so, you just put the alb target group arns from the load balancer module into the autoscaling module, with something like target_group_arns = module.alb.target_group_arns

Regardless of module, target_group_arns is a field on the aws_autoscaling_group terraform resource

Thanks, and sorry for the typo. Your info helped.

A follow up to yesterday’s question. I am using the CloudPosse ALB module (https://github.com/cloudposse/terraform-aws-alb) in conjunction with the CloudPosse ASG module (https://github.com/cloudposse/terraform-aws-ec2-autoscale-group). I linked them via target_group_arns as suggested yesterday. I instructed the ASG to used a standard, Linux AMI as its image, and I also tell the ASG to install httpd, etc. via userdata. However, I keep getting a 504 Gateway Time-out error. During troubleshooting, I noticed that the registered targets in my target group are failing their health check with 504 errors. When I look at the actual EC2 instances, they are using the default VPC security group which has no ingress or egress rules. So I found my 504 problem, but I’m not certain why my targets don’t have the proper security groups. The module is generating the expected security group to let in [0.0.0.0/0] over port 80, but that security group is not assigned to the targets in my target group. I see that the security groups are assigned to the ENIs, but that’s it. Any help/advice is most appreciated.

May have figured this out. I needed to get the security group from the ALB module and feed it into the ASG module.

Hi Guys,
I’ve added some additional parameters into terraform-aws-codebuild
https://github.com/cloudposse/terraform-aws-codebuild/pull/53
Can some one review pls.
Thnks

Just a quick one to add a missing output: https://github.com/cloudposse/terraform-aws-rds/pull/59

Hey @Maxim Mironenko (Cloud Posse) - any movement on https://github.com/cloudposse/terraform-aws-tfstate-backend/pull/43 ?

Hi All,
Can some suggest module for ElastiCache (REDIS)

Hi im trying to create a multiple subnets with terraform-aws-multi-az-subnets . However, since count is not allowed within modules, is there a way to use a single module and have some kind of iteration over the cidr lists to generate the subnets?

Hello there, I'd like to disable the creation of the s3 endpoint when using the EMR module: https://github.com/cloudposse/terraform-aws-emr-cluster/pull/14 -- I've already got an S3 endpoint managed somewhere else.

@Maxim Mironenko (Cloud Posse) ☝️

(Also we have #pr-reviews channel)

The bot and these commands are really nice!

Anyone have a module I can plugin to get RDS event logging for cloudwatch events pushed to cloudwatch logs+pager duty or similar destination. I saw an older cloudposse one, some promise. Anything else?

@Andriy Knysh (Cloud Posse) i have been using cloudposse for long time , hey just need a direction for how to include a provision for reading a another account bucket
i have been using private subnet for emr clusters

Hey folks - Opened a smol PR - https://github.com/cloudposse/terraform-aws-route53-alias/pull/21 was hoping for maybe a quick turnaround? 😄 cc @Erik Osterman (Cloud Posse) / @Maxim Mironenko (Cloud Posse)

@Andriy Knysh (Cloud Posse) can you review this ☝️

yes

set the channel topic: Terraform Modules

Thank you!

@joshmyers https://github.com/cloudposse/terraform-aws-route53-alias/releases/tag/0.6.0