Common issues

Attini CLI

Text ‘date’ could not be parsed at index 19

I get the error Text '2021-05-04T18:35:49UTC' could not be parsed at index 19 when using the CLI with AWS SSO credentials.

If you authenticate with an old version AWS CLI version this error occurs. Please update the AWS CLI.

For more information about this issue, see aws-sdk-java-v2/issues/2190.


Updating the Attini CLI doesn’t work

If you update the Attini CLI and the installation looks ok, but it takes no effect its probably because you have 2 Attini CLI installed and the “old” installation is located before your “new” installation on your system PATH.

Run the command: type attini

And verify that the Attini CLI path is actually the one you updated.


“bad CPU type in executable: attini” on Mac with Apple silicon processor

The Attini CLI is currently only compiled for x86_64 CPU architecture for Mac, so you need Rosetta 2 for it to work.

You can install Rosetta using the command:

/usr/sbin/softwareupdate --install-rosetta --agree-to-license

Note: The first time you run the Attini CLI after the Rosetta installation it will take a few extra seconds.


(zsh) Attini CLI broke after I updated it

If you get the error zsh: killed on all the Attini CLI commands after an update, you have to delete the Attini CLI and reinstall it. See the rm ${PATH_TO_CLI}/attini; part of the installation Attini CLI command.


Getting error “sh: syntax error near unexpected token `(‘” when turning on cli autocomplete

If you are trying to enable Attini CLI autocomplete via source <(attini generate-completion) command and it fails like this:

sh: syntax error near unexpected token `('

Run this command instead:

attini generate-completion > /tmp/attini-generate-completion; source /tmp/attini-generate-completion; rm -f /tmp/attini-generate-completion

(bash) Attini CLI autocompletion does not work

If the autocompletion for the CLI does not work make sure you are not running an old version of bash. Attini autocompletion required bash version 4 or later. To update bash on a Mac with homebrew run brew install bash.


Deployment plan

Deployment plan froze

A frozen deployment plan is a consequence of the Attini resources or AWS Service integrations not responding the underlying AWS StepFunction. This can be due to issues (for example throttling) with some underlying services, like AWS Lambda or Amazon SNS.

If you have a frozen deployment plan you should retry your deployment. The Attini framework will always cancel the old deployments before starting a new one so there is no risk for “zombie deployment”.

Importing an old CloudFormation stack into a deployment plan

If you have an old CloudFormation stack that you include into an Attini deployment plan and the first update of that CloudFormation stack fails, the Attini deployment will freeze. The workaround is to not do any updates to the CloudFormation stack the first run, that way the CloudFormation stack will not fail.

Attini::Deploy::Runner ClusterNotFoundException

Error message: Cluster not found

The deployment plan crashed at a AttiniRunnerJob step.

This can be because the EcsCluster in the Attini::Deploy::Runner does not exists in the current AWS Account and Region and you need to create it.

But if the EcsCluster is omitted, the error happens when there is no default cluster. Create a default cluster with the AWS CLI command:

aws ecs create-cluster