Timeout error during deployment of Xenial from MAAS 2.0

[Chris Gregan]

Build Version/Date: 0.7.7~bzr1182-0ubuntu1

Environment used for testing: Xenial/MAAS 2.0a2

Summary:
Deployment of Xenial fails due to timeout issues during cloud-init process

Steps to Reproduce:
1) Attempt machine deploy from MAAS 2.0a2
2) Watch console output

Expected result:
Deployed

Actual result:
Failure to deploy/Timeout erros

[Chris Gregan (cgregan)]

Difficult to get more info due to state of machine

[Scott Moser (smoser)]

The 'ping timeout' errors that you're seeing are open-iscsi complaining that the root volume is not reachable. ( https://sourceforge.net/p/iscsitarget/mailman/message/26138758/ for example).
That wasn't fatal though.

The fatal failure occurred earlier was '503 service unavailable' which surely would remind someone of bug 1547640.

Its possible that the ping timeout errors are coming on shutdown, possibly if the nic got brought down on shutdown (even though it is not supposed to).

[Chris Gregan (cgregan)]

https://pastebin.canonical.com/152034/
https://pastebin.canonical.com/152033/
https://pastebin.canonical.com/152026/

[James Falcon (falcojr)]

This is filed against an old and unsupported version of cloud-init. If this is still a problem, please file a new bug with updated logs from the current cloud-init version.

[James Falcon (falcojr)]

Tracked in Github Issues as https://github.com/canonical/cloud-init/issues/2632