Bug 1848483

Summary: [OSP16.1][Routed-provider-networks]Scheduler isn't working properly
Product: Red Hat OpenStack Reporter: Candido Campos <ccamposr>
Component: openstack-neutronAssignee: Bernard Cafarelli <bcafarel>
Status: CLOSED DEFERRED QA Contact: Eran Kuris <ekuris>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 16.1 (Train)CC: akatz, amuller, bcafarel, chrisw, scohen, tfreger
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2021-03-29 14:11:08 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:
Embargoed:
Bug Depends On:    
Bug Blocks: 1671811    

Description Candido Campos 2020-06-18 12:54:03 UTC
Description of problem:

Scheduler is creating a vms only in the first segment until some vm is created in other segment using --host option

Version-Release number of selected component (if applicable):

(overcloud) [stack@site-undercloud-0 scripts]$ openstack subnet list
+--------------------------------------+---------------------------------------------------+--------------------------------------+------------------+
| ID                                   | Name
                    | Network                              | Subnet
       |
+--------------------------------------+---------------------------------------------------+--------------------------------------+------------------+
| 211f0091-fbc8-4659-8861-499519aa9034 | HA subnet tenant
b53dc5fbdb164fb3a37f792313f36a9b |
988d78a4-5520-45ad-aee4-2e1fdbd3feb2 | 169.254.192.0/18 |
| 23fb5b21-706b-4c82-954f-9fa846bc3364 | subnet2_public
                    | 2fd259e2-014f-44f5-b875-bcbd3c4bfba0 |
10.0.20.0/24     |
| 793abf2d-ef3c-44b1-b889-5f73a239a873 | net10
                    | ec060fa8-eda2-4ae5-8d5c-f48c0de98796 |
10.10.0.0/16     |
| 7fa8f66b-acc6-419e-bdc8-5d56e16e5bf0 | subnet1_public
                    | 2fd259e2-014f-44f5-b875-bcbd3c4bfba0 |
10.0.10.0/24     |
| 96ac2d22-add6-483e-ad80-1e5f18f61d02 | subnet3_public
                    | 2fd259e2-014f-44f5-b875-bcbd3c4bfba0 |
10.0.30.0/24     |
+--------------------------------------+---------------------------------------------------+--------------------------------------+------------------+


(overcloud) [stack@site-undercloud-0 scripts]$ openstack server list 
+--------------------------------------+------+--------+--------------------+--------+--------+
| ID                                   | Name | Status | Networks           | Image  | Flavor |
+--------------------------------------+------+--------+--------------------+--------+--------+
| 8bc873fe-1374-49bf-8c25-6e330b7f9ba2 | vm6  | ERROR  |                    | cirros |        |
| eca11825-7540-435a-986f-e25ab79128de | vm5  | ACTIVE | public=10.0.10.220 | cirros |        |
| 5d0e5185-478c-413d-835b-360a0647b7fd | vm4  | ACTIVE | public=10.0.10.54  | cirros |        |
| 8b8d093c-eb6c-4310-8629-add696b8c1bb | vm3  | ACTIVE | public=10.0.10.248 | cirros |        |
| 16d019f6-b6ab-44ad-a0f9-ccdc596ad45e | vm2  | ACTIVE | public=10.0.10.91  | cirros |        |
| e13dfd2f-3b70-4d57-bde8-35ef493c2ab2 | vm1  | ACTIVE | public=10.0.10.46  | cirros |        |
+--------------------------------------+------+--------+--------------------+--------+--------+

Comment 2 Bernard Cafarelli 2020-06-19 13:47:10 UTC
There is no specific indication on how placement, nova and neutron segments will end up distributing new VMs among segments - we will test with reduced compute nodes to see if deployment switches properly to other segments if one compute node/segment is full.

Note this is only in the case where we let placement decide which segment to use, and not force a specific one/a nova AZ

Comment 5 Alex Katz 2021-03-29 14:11:08 UTC
There is a documented limitation for this issue in OSP16.1