Open Stack

Hi,

That's quite an old spec. :)

It has quite a bit of history, and the general nova core opinions were "No, we're probably not going to do that" in the beginning, to "We should probably add that, if people are asking for it" during the last OpenStack PTG. It even had 2x +2s and ~11 +1s at one point.

I'll repropose the blueprint again for Stein and rebase everything, if people want it. Implementation-wise, it's pretty much done, last time I was still doing some functional and unit tests for it. I had a tempest test for it too. I had some issues regarding notifications, but gibi helped me sort them out (thanks!).

As far as functionality goes, I'm afraid that at the moment we're only going to add live-resize to a larger flavor (live-upsizing). There are a few concerns regarding live-downsizing, especially when it comes to the hypevisor support for something like this (not all of them supports this). Additionally, when it comes to live-downsizing the disk, AFAIK, there's also a data loss concern associated with it, if the disk was not freed and / or if the partition table wasn't shrunk beforehandm, etc. Additionally, some nova drivers don't support downsizing the disks at all, not even cold resize.

There are a lot of discussions and debates on the spec you've linked, and you can also read the arguments regarding the question you've asked too.

tl;dr version. During the last PTG, we've agreed to "approve" the blueprint with only live-upsizing the instance in-place as the first iteration of the feature, which any new additions to be discussed afterwards.

Best regards,

Claudiu Belu


________________________________
From: Rambo [lijie at unitedstack.com]
Sent: Monday, August 13, 2018 2:36 PM
To: OpenStack Developmen
Subject: [openstack-dev] [Openstack-operators][nova]about live-resize down the instance

Hi,all

      I find it is important that live-resize the instance in production environment,especially live downsize the disk.And we have talked it many years.But I don't know why the bp[1] didn't approved.Can you tell me more about this ?Thank you very much.

[1]https://review.openstack.org/#/c/141219/








Best Regards
Rambo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20180813/f4c6cdc2/attachment-0001.html>