Bug 1321005

Summary: [UX] Deployment should fail right at the beginning if not enough hosts are available
Product: Red Hat OpenStack Reporter: Udi Kalifon <ukalifon>
Component: rhosp-directorAssignee: Angus Thomas <athomas>
Status: CLOSED NOTABUG QA Contact: Arik Chernetsky <achernet>
Severity: high Docs Contact:
Priority: unspecified    
Version: 8.0 (Liberty)CC: dbecker, dtantsur, jcoufal, mburns, morazi, rhel-osp-director-maint
Target Milestone: ---   
Target Release: 10.0 (Newton)   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2016-10-14 14:43:14 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Udi Kalifon 2016-03-24 13:01:15 UTC
Description of problem:
I designated 2 ironic nodes to be for compute-0 and compute-1:

ironic node-update d0fd91fb-c66e-4a5d-9b66-e3806959b31c replace properties/capabilities='node:compute-0,boot_option:local'
ironic node-update 3e097f7b-7dd1-47de-b4a4-345248510b6a replace properties/capabilities='node:compute-1,boot_option:local'

I added scheduling hints to my environment file:

parameter_defaults:
  NovaComputeSchedulerHints:
    'capabilities:node': 'compute-%index%'

I then tried to deploy with 3 computes. It takes a really long time for the deployment to fail. Eventually it fails with "not enough hosts available".


Version-Release number of selected component (if applicable):
python-tripleoclient-0.3.1-1.el7ost.noarch
openstack-tripleo-heat-templates-kilo-0.8.12-2.el7ost.noarch


How reproducible:
100%


Steps to Reproduce:
1. Set ironic attributes on several nodes, and assign them to the compute(s) deployment
2. Try to deploy with more computes than you have allocated.


Actual results:
Deployment fails as it should, but this error should stop you right from the start... No need to waste time.

Comment 2 Mike Burns 2016-04-07 21:36:02 UTC
This bug did not make the OSP 8.0 release.  It is being deferred to OSP 10.

Comment 3 Dmitry Tantsur 2016-10-14 14:43:14 UTC
Hi! We can't really replicate the whole scheduling that Nova does in TripleO, it is way too complex. We do check for at least minimum number of nodes in the "available" state, and we do check profiles. I don't think we can do anything else, sorry.