Fuel for OpenStack

[astute] Scheduler doesn't get appropriate info about nodes status

Bug #1269460 reported by Egor Kotko
10
This bug affects 2 people
Affects Status Importance Assigned to Milestone
Fuel for OpenStack
Confirmed
Medium
Vladimir Sharshov
Fuel for OpenStack 6.1

Bug Description

{"build_id": "2014-01-15_01-17-41", "ostf_sha": "f4795bbea783105f22834746cb20ff3578c116fe", "build_number": "33", "nailgun_sha": "ccef1e2e529d2f6789a19b3e204deabd5fdb2545", "fuelmain_sha": "e65f738c4a76967ea521a1b6031344816a618906", "astute_sha": "ca787b5b0a3a418e6885b9fd2d795c9fd158ed0a", "release": "4.1", "fuellib_sha": "c8673bb9474ccb0a51fb9077910b009ff2d9034b"}

Steps to reproduce:
1. Create instance.
2. Delete 1 compute node.
3. Migrate instance.
Configuration of cluster:
Ubuntu simple, 1Controller, 3 (compute+ceph), Ceph configuration - images, nova + FlatDHCP, Lanch instance from image.

Expected result:
Compute node must be deleted from openstack database

Actual result:
Deleted compute node present in database with active status.
The system allows to migrate instances on deleted compute node.

Kun Huang (academicgareth)
Changed in nova:
assignee: nobody → Kun Huang (academicgareth)
Revision history for this message
Kun Huang (academicgareth) wrote :

How did you delete the compute node?

Revision history for this message
Egor Kotko (ykotko) wrote :

Via Fuel.

Kun Huang (academicgareth)
affects: nova → fuel
Kun Huang (academicgareth)
Changed in fuel:
assignee: Kun Huang (academicgareth) → nobody
Tatyanka (tatyana-leontovich)
Changed in fuel:
milestone: none → 4.1
Egor Kotko (ykotko)
tags: added: nova
Dmitry Pyzhov (dpyzhov)
Changed in fuel:
assignee: nobody → Fuel Library Team (fuel-library)
importance: Undecided → Low
Dmitry Pyzhov (dpyzhov)
Changed in fuel:
assignee: Fuel Library Team (fuel-library) → Fuel Astute Team (fuel-astute)
importance: Low → Medium
status: New → Confirmed
tags: added: astute
removed: nova
Revision history for this message
Vladimir Kozhukalov (kozhukalov) wrote :

We need to implement removing node from nova before we actually delete (clear node's hard drive) node and reboot it. We can add this piece of code into astute using "fog" (http://fog.io/) or any other ruby binding to openstack.

Dmitry Pyzhov (dpyzhov)
Changed in fuel:
milestone: 4.1 → 5.0
Revision history for this message
Dima Shulyak (dshulyak) wrote :

astute will need controller ip address from management network, so it will require adding it in delete task message generator in nailgun.
also astute can use proxy that is used for ostf

Dmitry Pyzhov (dpyzhov)
Changed in fuel:
milestone: 5.0 → 5.1
Vladimir Sharshov (vsharshov)
Changed in fuel:
assignee: Fuel Astute Team (fuel-astute) → Vladimir Sharshov (vsharshov)
Dmitry Ilyin (idv1985)
summary: - Scheduler doesn't get appropriate info about nodes status
+ [astute] Scheduler doesn't get appropriate info about nodes status
Mike Scherbakov (mihgen)
Changed in fuel:
milestone: 5.1 → 6.0
Revision history for this message
Vladimir Sharshov (vsharshov) wrote :

This tasks looks like pre/post deployment actions in case of 'remove'. I believe that i will be good to release this after we implement 'granular deployment' (~ 6.1) or using Nailgun plugins mechanizm (6.0). More update will be after implementation of Nailgun plugins

Revision history for this message
Roman Prykhodchenko (romcheg) wrote :

It hits code freeze in 6.0. Moving to 6.1

Changed in fuel:
milestone: 6.0 → 6.1
Revision history for this message
Christopher Aedo (docaedo) wrote :

This patch added the API method for removing the compute node:
https://review.openstack.org/#/c/39998/

If a new ruby binding is necessary, Aviator should be considered as a better alternative to fog: https://github.com/aviator/aviator

Mike Scherbakov (mihgen)
tags: added: operations
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Related blueprints

Remote bug watches

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