Bug 1396107 - [RFE] Nova should try to stop instances gracefully [NEEDINFO]
Summary: [RFE] Nova should try to stop instances gracefully
Keywords:
Status: CLOSED INSUFFICIENT_DATA
Alias: None
Product: Red Hat OpenStack
Classification: Red Hat
Component: openstack-nova
Version: 8.0 (Liberty)
Hardware: x86_64
OS: Linux
high
high
Target Milestone: async
: ---
Assignee: Eoghan Glynn
QA Contact: Prasanth Anbalagan
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2016-11-17 12:55 UTC by Ondrej
Modified: 2021-02-01 02:41 UTC (History)
11 users (show)

Fixed In Version:
Doc Type: Enhancement
Doc Text:
Clone Of:
Environment:
Last Closed: 2017-04-27 17:43:46 UTC
Target Upstream Version:
rcernin: needinfo? (eglynn)


Attachments (Terms of Use)

Description Ondrej 2016-11-17 12:55:18 UTC
Description of problem:
Hello,

When executing a "nova stop" command, instances do not have the option to graceful shutdown the OS.

For some instances (Windows mostly), this behavior is causing many problems. Is there any option to implement a new feature to have both options:

    1- Nova stop --> should try a graceful shutdown (maybe a --timeout parameter or nova stop-soft?)
    2-Nova force stop --> current behavior


This is an important feature because many instances has been affected. In customer upgrade procedure admins are stopping all instances performing a "nova stop" command. When we start instances again, some of them (Windows mostly) are corrupted.


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


How reproducible:
always

Steps to Reproduce:
1.
2.
3.

Actual results:


Expected results:
Being able to gracefully shut the os down with nova stop

Additional info:

Comment 2 Stephen Gordon 2016-11-25 16:55:25 UTC
Related: https://blueprints.launchpad.net/nova/+spec/user-defined-shutdown

Comment 3 Stephen Gordon 2016-11-25 17:03:29 UTC
(In reply to Stephen Gordon from comment #2)
> Related: https://blueprints.launchpad.net/nova/+spec/user-defined-shutdown

This was implemented some releases ago now, and attempts to perform a graceful shutdown by default, are you sure the issue isn't that the guests in question take a long time to shutdown and as a result are forced off after the timeout expires?

The default timeout is 60 seconds as defined by the shutdown_timeout config setting. It can also be set per guest using the image_os_shutdown_timeout image property.

Comment 4 Red Hat Bugzilla Rules Engine 2017-04-27 12:11:26 UTC
This bugzilla has been removed from the release and needs to be reviewed and Triaged for another Target Release.


Note You need to log in before you can comment on or make changes to this bug.