Bug 1899468

Summary: [RFE][OSP 18] Provide a workaround option and/or nova-manage command to force the refresh of connection_info during a hard reboot
Product: Red Hat OpenStack Reporter: Lee Yarwood <lyarwood>
Component: openstack-novaAssignee: Lee Yarwood <lyarwood>
Status: CLOSED CURRENTRELEASE QA Contact: OSP DFG:Compute <osp-dfg-compute>
Severity: high Docs Contact:
Priority: urgent    
Version: 17.0 (Wallaby)CC: dasmith, eglynn, gcharot, jhakimra, jmelvin, kchamart, sbauza, sgordon, vromanso
Target Milestone: ---Keywords: FutureFeature, Triaged
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of:
: 1999582 (view as bug list) Environment:
Last Closed: 2022-01-24 09:55:19 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: Xena
Embargoed:
Bug Depends On:    
Bug Blocks: 1999582, 1999583, 1999588    

Description Lee Yarwood 2020-11-19 10:43:17 UTC 
Description of problem:
Provide a workaround option and/or nova-manage command to force the refresh of connection_info during a hard reboot.

This is useful in the context of live migration rollback failures, environmental changes and other future issues where the fact that Nova caches connection_info for a given volume from Cinder can cause issues.

A workaround option and/or nova-manage command would allow operators to resolve these issues without requiring code changes or a cold migration to another host.

This should only be possible when the instance is SHUTOFF or when hard rebooting the instance.

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

How reproducible:
Always.

Steps to Reproduce:
1. Cause stale connection_info to be populated for a given block_device_mapping in the Nova database.

Actual results:
There is no way to force a refresh of this stale connection_info without code changes or a cold migration to another host.

Expected results:
An operator should be able to force a refresh of the connection_info while the associated instance is SHUTOFF.

Additional info: