Maverick: cloud-init fails during boot of instance

[Torsten Spindler]

Binary package hint: cloud-init

I get an error from cloud-init fairly frequently (approx 1 out of 10 launches). The traceback looks like

Traceback (most recent call last):
  File "/usr/bin/cloud-init", line 90, in <module>
    main()
  File "/usr/bin/cloud-init", line 47, in main
    cloud.consume_userdata,[],False)
  File "/usr/lib/python2.6/dist-packages/cloudinit/__init__.py", line 215, in se
m_and_run
    if self.sem_has_run(semname,freq): return
  File "/usr/lib/python2.6/dist-packages/cloudinit/__init__.py", line 173, in se
m_has_run
    semfile = self.sem_getpath(name,freq)
  File "/usr/lib/python2.6/dist-packages/cloudinit/__init__.py", line 167, in se
m_getpath
    freqtok = self.datasource.get_instance_id()
  File "/usr/lib/python2.6/dist-packages/cloudinit/DataSourceEc2.py", line 64, i
n get_instance_id
    return(self.metadata['instance-id'])
KeyError: 'instance-id'
init: cloud-init main process (567) terminated with status 1

More complete (different) log on pastebin, http://pastebin.ubuntu.com/525732/

This happens on Maverick with eucalyptus 2.0+bzr1241-0ubuntu4 on the server and on the lucid instance cloud-init 0.5.10-0ubuntu1

[Scott Moser (smoser)]

I really think this is a duplicate of bug 571271, which is fix-released in lucid-updates.
The problem is that the metadata service is not completely up when cloud-init hits it. It is really a Eucalyptus bug, but is worked around by cloud-init by waiting longer (1050 seconds) before giving up.