This only occurred once in todays beta-1 testing. I have seen it on
very rare occasion, as noted above.
I started digging a bit through my build logs. The daily build process
launches an m1.large instance in each region to facilitate populating
EBS root instances. On 2011-03-03, I updated the utility instances that
we use to the latest lucid refreshed version (ebs/ubuntu-
lucid-10.04-amd64-server-20110201.1), so this first info represents data
only from that date.
- 66 build logs (17 failed for one reason or another)
- 298 instances launched from 2011-03-03 to 2011-03-30.
- 9 of 298 instances could not be reached via ssh
- 3 of those 9 instances were due to timeout on metadata service these were:
2011-03-26 us-west-1 maverick-desktop-uec-20110326
2011-03-28 us-west-1 natty-server-uec-20110328
2011-03-29 us-west-1 lucid-server-uec-20110329
So, 298 instances launched, 3 failures on metadata service. All of
those in us-west-1 and occurred in the last 5 days.
Going back further, though all the logs I have since ~ 2010-03-16
- 704 build logs
- 2620 instances launched
- 28 failed to have elastic IP associated and be reached via ssh.
- 4 TOTAL have messages in console suggesting time out on metadata service
2011-03-01 us-east-1 lucid-server-uec-20110301
2011-03-26 us-west-1 maverick-desktop-uec-20110326
2011-03-28 us-west-1 natty-server-uec-20110328
2011-03-29 us-west-1 lucid-server-uec-20110329
** Changed in: cloud-init (Ubuntu)
Status: Incomplete => New
** Changed in: cloud-init (Ubuntu)
Importance: Undecided => Low
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to cloud-init in Ubuntu.
https://bugs.launchpad.net/bugs/745930
Title:
cloud-init timeout waiting for metadata service on EC2
--
Ubuntu-server-bugs mailing list
Ubuntu-server-bugs@lists.ubuntu.com
Modify settings or unsubscribe at:
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server-bugs
