Comment 1 for bug 1017418

Is it possible that your script is deleting instances before they are done creating? I don't see any way for the code path with the error to get hit unless the network deallocate is getting called before the allocation is done somehow. Can you try either waiting until the api says the instance is running or upping the wait to a minute to see if the error persists? That will help us track down if it is a race condition or a logic error somewhere.

Vish