Fails to delete stack if extension no longer exists

Bug #1641249 reported by Turbo Fredriksson
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
OpenStack Heat
New
Low
Unassigned

Bug Description

After my (accidental) upgrade from Mitaka to Newton a few weeks ago (I've been busy with work, so I haven't had time to deal with all the issues I got because of this until now), some of my services no longer exists (or haven't been correctly configured).

So when I try to delete my stack(s), I get things like:

HEAT-E99001 Service neutron is not available for resource type OS::Neutron::FirewallPolicy, reason: Required extension fwaas in neutron service is not available.

HEAT-E99001 Service neutron is not available for resource type OS::Neutron::PoolMember, reason: Required extension lbaas in neutron service is not available.

ERROR: Resource floating description type must be string

Revision history for this message
Thomas Herve (therve) wrote :

It really feels like an operation problem. We may be able to do something when deleting stacks, but what about updating them, etc?

Changed in heat:
importance: Undecided → Low
milestone: none → next
Revision history for this message
Turbo Fredriksson (turbo-bayour) wrote :

Updating a "broken" stack should always fail, shouldn't it? I mean, it can't update something when the underlying resources don't exist [any longer]..

Unless it's possible to create a stack with, for example, a LBaaSv1. Then modify the stack 'source' with LBaaSv2 and have heat automatically remove the LBaaSv1 part and 'only' add the LBaaSv2 part..

Feel free to try to solve that :). But I'm happy with a fail in that case and as long as I get an option to now delete my "undeletable" stacks, I'll be VERY happy!

I "can't" create my new stacks without having cleaned up the old ones. Heat managed to delete almost all of their resources, but there's a few bits and pieces in them that can't, currently, be deleted and currently that's all _I_ want.

But having Heat do the transition would probably be "a good thing" for the long run..

Revision history for this message
Rabi Mishra (rabi) wrote :

I think it's risky to allow deleting a stack when it contains resources from a service that is down(can be back up after sometime) and would leave resources managed by heat abandoned.

As Thomas mentioned there probably can be force delete option to delete stacks irrespective of availability of services.

Revision history for this message
Turbo Fredriksson (turbo-bayour) wrote :

I'm good with that.

Revision history for this message
Turbo Fredriksson (turbo-bayour) wrote :

Is there any news on this? It's currently blocking my setup..

Revision history for this message
Turbo Fredriksson (turbo-bayour) wrote :

A workaround would be sufficient at this point.

How bad would deleting the 'heat' database and then create it again be? Can/should I just delete entries 'here and there'?

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.