Elastic Beanstalk stuck for more than 12 hrs

Question

I have an EB application that is stuck with this error: Failed Environment update activity. Reason: Internal Failure. I can not abort anything and I cannot change any configuration.. I dont know what to do, because every thing I do, it says: Could not abort the current environment operation for environment-xxxxxx: Environment named environment-xxxx is in an invalid state for this operation. Must be pending deployment.

I want to destroy the application or remove everything to start again, but I cannot do that.

Does anyone know how to unlock this?

Thanks in advance!

Solution

If you know how to fix the source of problem (for example: change application settings in Beanstalk Environment variables or deploy fixed version of your application), then:

1. Go to page Auto Scaling Group, choose your region, find Auto Scaling Group by Beanstalk Environment ID (like e-abcd12345).
2. In Details tab, push "edit", set Desired, Min and Max to 0. Remember previous settings.
3. In Beanstalk page, wait for message like "Removed instances [i-xxx, i-yyy] from your environment."
4. Change settings or deploy your application
5. Return Autoscaling settings back.
6. Be patient, it may took 5-10 minutes to change Beanstalk health to "Ok".

Comments

[kumar kundan]

thanks.. you made my day

[sagar junnarkar]

This answer saved my job. ;) Thanks.

[carlin.scott]

I had this happen in my production environment, so setting desired capacity to 0 wasn't an option. Instead I terminated the new instance created by the deployment from the Monitoring tab, and then I followed your directions but set min and desired capacity the pre-deployment values. After about 8 minutes, EB recovered from the stuck state like you said it would. Thank you!