instance went straight to terminated

[Scott Moser]

When running instances for the iso tests for maverick alpha 1 (20100601) I had one instance go directly to terminated:

INSTANCE i-e18ec0b3 ami-7de29d2f terminated ec2-keypair.ap-southeast-1 6 m1.large 2010-06-01T23:27:53+0000 ap-southeast-1b aki-6bef9039 monitoring-disabled instance-store

There was no console output.

ProblemType: Bug
DistroRelease: Ubuntu 10.10
Package: linux-ec2 2.6.32.305.6
ProcVersionSignature: User Name 2.6.32-305.9-ec2 2.6.32.11+drm33.2
Uname: Linux 2.6.32-305-ec2 x86_64
Architecture: amd64
Date: Wed Jun 2 13:06:03 2010
Ec2AMI: ami-7de29d2f
Ec2AMIManifest: ubuntu-images-testing-ap-southeast-1/ubuntu-maverick-daily-amd64-server-20100601.manifest.xml
Ec2AvailabilityZone: ap-southeast-1b
Ec2InstanceType: m1.large
Ec2Kernel: aki-6bef9039
Ec2Ramdisk: unavailable
ProcEnviron:
 LANG=en_US.UTF-8
 SHELL=/bin/bash
SourcePackage: linux-meta-ec2

[Scott Moser (smoser)]

found this during maverick alpha2 testing:
IMAGE ami-ba7991d3 099720109477/ubuntu-images-testing/ubuntu-maverick-daily-i386-server-20100630.1 099720109477 available public i386 machine aki-078e676e instance-store

That kernel is : ubuntu-maverick-i386-linux-image-2.6.32-305-ec2-v-2.6.32-305.9-kernel.

[Eric Hammond (esh)]

I suspect that this may possibly be caused by an occasional (rare) EC2 glitch. If it doesn't happen regularly to multiple people, then it may not be worth investigating. EC2 instances are not infallible and users should be designing their architectures to survive the occasional failed instance.

I used to see straight to terminated more from EC2 a couple years back, but the incidence has dramatically reduced to near zero.