Scott's Weblog The weblog of an IT pro focusing on cloud computing, Kubernetes, Linux, containers, and networking

Using the AWS CLI to Tag Groups of AWS Resources

To conduct some testing, I recently needed to spin up a group of Kubernetes clusters on AWS. Generally speaking, my “weapon of choice” for something like this is Cluster API (CAPI) with the AWS provider. Normally this would be enormously simple. In this particular case—for reasons that I won’t bother going into here—I needed to spin up all these clusters in a single VPC. This presents a problem for the Cluster API Provider for AWS (CAPA), as it currently doesn’t add some required tags to existing AWS infrastructure (see this issue). The fix is to add the tags manually, so in this post I’ll share how I used the AWS CLI to add the necessary tags.

Without the necessary tags, the AWS cloud provider—which is responsible for the integration that creates Elastic Load Balancers (ELBs) in response to the creation of a Service of type LoadBalancer, for example— won’t work properly. Specifically, the following tags are needed:

kubernetes.io/cluster/<cluster-name>
kubernetes.io/role/elb
kubernetes.io/role/internal-elb

The latter two tags are mutually exclusive: the former should be assigned to public subnets to tell the AWS cloud provider where to place public-facing ELBs, while the latter is assigned to private subnets to manage the placement of internal ELBs.

When CAPA is managing the infrastructure, this isn’t a problem because CAPA will add the necessary tags when it creates the infrastructure. Therefore, had I been able to use a separate VPC for each cluster, I could have let CAPA manage the infrastructure and avoided any issues entirely. In this case I was using a separate infrastructure-as-code tool (Pulumi) to manage the underlying AWS infrastructure and had the requirement to use a single VPC for multiple clusters.

Now, I could have logged into the AWS console and used “point-and-click” to work my way through tagging the VPC and the subnets, but I preferred to use the AWS CLI. I quickly found the aws ec2 create-tags command, which would do exactly what I needed; all I had to do was provide a list of the resources to tag and the tags to add.

To find the resources, I just had to make use of the tags that I’d made sure to assign to all the resources I created. So, to find all my public subnets, I used this AWS CLI command:

aws ec2 describe-subnets --filters Name=tag:Owner,Values="Scott Lowe" Name=tag:Name,Values="*pub*" --query 'Subnets[*].SubnetId' --output text

Similarly, I could pull up my private subnets like this:

aws ec2 describe-subnets --filters Name=tag:Owner,Values="Scott Lowe" Name=tag:Name,Values="*priv*" --query 'Subnets[*].SubnetId' --output text

Next, I had to prepare the tags I wanted added to each resource. For this, the parameter --generate-cli-skeleton input was very helpful; it generated the following framework:

{
    "DryRun": true,
    "Resources": [
        ""
    ],
    "Tags": [
        {
            "Key": "",
            "Value": ""
        }
    ]
}

Using this skeleton as the foundation, I created two JSON input files—one for the tags to be assigned to public subnets, and one for the tags to be assigned to private subnets.

The documentation for the aws ec2 create-tags command indicated that it would take a space-delimited list of resource IDs, and the output from the aws ec2 describe-subnets command appeared to be space-delimited. At this point, I thought I’d be able to do something like this:

aws ec2 describe-subnets --filters Name=tag:Owner,Values="Scott Lowe" \
Name=tag:Name,Values="*priv*" --query 'Subnets[*].SubnetId' \
--output text | xargs -I {} aws ec2 create-tags --resources {} \
--cli-input-json file://priv-subnet-tags.json

Alas, this did not work. Further, no amount of messing around with the output of the aws ec2 describe-subnets command could get it into a format that aws ec2 create-tags liked. I’m sure it was an error of some sort on my part, but I couldn’t figure it out.

I’d been spending a fair amount of time with jq recently (parsing a lot of Envoy configurations), so I thought, “Why not drop back to JSON output and use jq?”

Dropping the --output text and piping output through jq finally got me to a working command. Here’s the command for the public subnets:

aws ec2 describe-subnets --filters Name=tag:Owner,Values="Scott Lowe" \
Name=tag:Name,Values="*pub*" --query 'Subnets[*].SubnetId' \
jq -r '.[]' | xargs -p -I {} -n 1 aws ec2 create-tags --resources {} \
--cli-input-json file://pub-subnet-tags.json

There was also a corresponding version for the private subnets as well.

Along the way, I did end up accidentally applying some private tags to public subnets; fortunately, the aws ec2 delete-tags command was there to save me.

Once the (correct) tags were applied, I was able to create a series of workload clusters via CAPI, and everything worked just as expected.

What did I learn from this whole process?

  • I had not been previously aware of the aws ec2 create-tags and aws ec2 delete-tags commands; these are pretty handy.
  • It seems that working with structured data, like JSON, can sometimes be easier than freeform text. jq is your ally here.
  • Using the --generate-cli-skeleton input parameter is very useful for generating JSON input documents. I’ll definitely be using that one again.

I hope this information is useful to you in some way. Thanks for reading! Feedback is always welcome, so feel free to reach out to me on Twitter if you have any questions or comments.

Metadata and Navigation

Be social and share this post!