Dashboard already exists

[MarkVink]

This is a Bug Report

Description

Deployment fails due to dashboard already existing (created during a previous deployment).

For bug reports:

• What went wrong? -> Serverless deployment fails because the dashboard was created in a previous run.

• What did you expect should have happened? -> I expect the deployment to succeed

• What was the config you used? -> Only turning on the dashboard

custom: {
  <snip>
  alerts: {
    dashboards: true,
  },
},

• What stacktrace or error message from your provider did you see?

✖ Stack <snip>-staging failed to deploy (114s)
Environment: darwin, node 18.12.1, framework 3.27.0 (local), plugin 6.2.3, SDK 4.3.2
Credentials: Local, "<snip>" profile
Docs:        docs.serverless.com
Support:     forum.serverless.com
Bugs:        github.com/serverless/serverless/issues

Error:
CREATE_FAILED: AlertsDashboard (AWS::CloudWatch::Dashboard)
Dashboard '<snip>-staging-us-east-1' already exists

Additional Data

• Serverless Framework Version: 3.27.0
• Stack Trace:
• Provider Error messages:

[georgeoffley]

👍🏽 on this.

I was seeing the same thing when deploying. I had to go in and delete the dashboard manually and then rerun CI deployed.

I also noticed the same thing with our custom alarms: the same error message (Alarm already exists) and the same method for clearing the error.