Juju ignores vpc for cleanup operations

[Nicholas Skaggs]

Attempting to invoke juju kill-controller -t 0 or juju remove-machine to directly invoke the AWS api doesn't work. The log reads:

ERROR juju.worker.dependency engine.go:546 "undertaker" manifold worker returned unexpected error: destroying instances: No default VPC for this user (VPCIdNotSpecified)

However, the controller was bootstrapped with this information, and the controller config contains the correct VPC.

juju bootstrap aws aws --config vpc-id=vpc-12345 --config vpc-id-force=true

[John A Meinel (jameinel)]

I'm pretty sure we do the right thing around normal "juju
destroy-controller", but probably "juju kill-controller" running from the
client isn't tracking VPC correctly. Certainly need some testing on regions
with multiple VPCs, and ensuring we don't kill the wrong machines, etc.

On Wed, Oct 4, 2017 at 2:20 AM, Anastasia <email address hidden>
wrote:

> ** Tags added: network
>
> --
> You received this bug notification because you are subscribed to juju.
> Matching subscriptions: juju bugs
> https://bugs.launchpad.net/bugs/1718028
>
> Title:
> Juju ignores vpc for cleanup operations
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/juju/+bug/1718028/+subscriptions
>

[Canonical Juju QA Bot (juju-qa-bot)]

This bug has not been updated in 5 years, so we're marking it Expired. If you believe this is incorrect, please update the status.