SSH into EC2 instance failed

Question

I deployed by EC2 instance long back in Nov 2023 but all of sudden today morning it was neither reachable via HTTPS nor SSH. Nothing was touched on security group settings

Looking at the monitoring logs. I saw CPU utilization went up to 99% and N/W IN and OUT spiked considerably to ~5K account for 18MB request

I am using T2 Micro. Can this n/w request lead to EC2 unresponsive?

Restarting the EC2 instance fixed it but still a concern on what caused this high inbound request

My hunch is this could be  a DDOS but has anyone else faced it too?

Answer

Your logs would be the best place to look for the inbound traffic that you have noticed. If this is a web-server, look in the web-server logs for the time that the unusually high traffic was experienced.

Also, bear in mind that a t2.micro instance type is a burstable instance type. This has a baseline CPU performance of 10%. Anything over that, you are consuming CPU credits. A t2.micro earns 6 CPU credits an hour. Each credit is equal to a vCPU running at 100% for 1 minutes, that you get 6 minutes of burst up to 100% every hour, and it can earn a maximum of 144 credits. 6 credits an hour corresponds to a t2.micro's baseline CPU utilization of 10%.

So from this  you can appreciate that if you your instance is running at less than 10% it will accumulate credits and more than 10% it uses them.

Once your credits are exhausted - your CPU can no longer burst beyond the baseline.

High traffic for an extended period can therefore cause your instance to consume any accumulated CPU credits, after which it will get to only run at 10% CPU utilization.

This may be the cause of the inability to login to the instance.

SSH into EC2 instance failed

Relevant content