My EC2 tg4.micro simply disappeared today with no trace

Well that's terrifying! I don't have an answer sadly. Out of curiosity, was it a spot instance?

Me too =/ ...Never thought that this kind of thing could happen with no trace...

Yup, it was a spot instance, but even if it was the case that my credits expire or reached some kind of limit, my instance should be stopped but never deleted and I have no policy/role for that so...I even lookup Spot Instance pricing history for any changes but its completely stable for t4gmicro in that region...

I already opened a support case because I still have hope that inside AWS they can find what happen reading logs that customers don't have access to.

brunof wrote:
Me too =/ ...Never thought that this kind of thing could happen with no trace...

Yup, it was a spot instance, but even if it was the case that my credits expire or reached some kind of limit, my instance should be stopped but never deleted and I have no policy/role for that so...I even lookup Spot Instance pricing history for any changes but its completely stable for t4gmicro in that region...

I already opened a support case because I still have hope that inside AWS they can find what happen reading logs that customers don't have access to.
Hello, just an update here.

It really was an EC2 Spot related cause, when I though interrupted I didn't assimilate it as erased =P

I just received a helpful reply from AWS support.

So I probably set a “maximum Spot price" when creating the instance, and it was claimed and terminated after a 2-minute notification(which I thought was an email).

See more:
You can read more about Spot Instance requests and terminations in the "Understanding Spot Instances and Spot Price" section of our Spot Instances page:

http://aws.amazon.com/ec2/spot-instances

The EBS cause was also explained by the AWS:
When a Spot instance is terminated:

1) All data on instance-storage will be lost.
2) All EBS volumes that are attached which are set to "Delete on Terminate" will be deleted.
3) All EBS volumes that are attached which are set to NOT "Delete on Terminate" will be left unattached to any instance. You can then attach them to another instance and get to your data.

So, my mistake folks, still faithful using AWS again ;)

brunof wrote:
Me too =/ ...Never thought that this kind of thing could happen with no trace...

Yup, it was a spot instance, but even if it was the case that my credits expire or reached some kind of limit, my instance should be stopped but never deleted and I have no policy/role for that so...I even lookup Spot Instance pricing history for any changes but its completely stable for t4gmicro in that region...

I already opened a support case because I still have hope that inside AWS they can find what happen reading logs that customers don't have access to.
Hello, final update here.

Yesterday I found the log at CloudTrails -> Event history with one entry with name BidEvictedEvent

By default the behavior is to terminate the Spot instance, checked on cli with:
aws ec2 describe-spot-instance-requests

Useful link:
[Why is my Spot Instance terminating even though the Spot price doesn't exceed the maximum price?]:
https://aws.amazon.com/pt/premiumsupport/knowledge-center/ec2-spot-instance-termination-reasons/

[Specify the interruption behavior]
https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/spot-interruptions.html#specifying-spot-interruption-behavior

Edited by: brunof on Aug 6, 2021 6:28 AM