I've had an Elastic Beanstalk (Docker running on 64bit Amazon Linux 2023/4.1.0) instance running for around 10 weeks with no problems. When I pushed an update on Sunday using the EB CLI, my entire instance irrecoverably crashed and I got the error
Auto Scaling activity failed XX minutes ago with error: The parameter SecurityGroup is not recognized. Launching EC2 instance failed. (I see this error every time it tries to restart with a different number of minutes)
Google searches indicate that 3-4 other people have experienced this, but there are no answers on SO or Reddit that are helpful. I don't think I changed anything in the configuration, just pushed a few standard bugfixes and additional features to my Docker image. I've even resorted to ChatGPT for help, but while it can identify the vague area, all the suggestions are too general to be helpful.
While I am an experienced developer, I am quite new to AWS. Please feel free to explain anything like you would to a small child, or a dog familiar with Docker and Python.
I've tried cloning the environment, redeploying and examining the security groups. The former has failed every time, and when I try to redeploy I'm told that the environment is not in a state to accept the redeployment.
I don't know where to start with the Security Groups, and all the documentation I found is about creating and/or modifying them, not that the parameter itself is not recognised. I have 3 security groups, and when I try and use them all, I'm told that the 'character limit' is exceeded.
Might be not related to above issue but I was getting exactly the same error message when creating new Elastic Beanstalk on Java.
The issue was in Virtual Private Cloud (VPC). I wasn't selecting any with assumption that it will use default one. After I selected the default VPC explicitly, the issue is gone.
I gave a name to my default VPC in VPC section but not sure if it is related.