On 06/28/2013 05:58 PM, John Garbutt wrote:
I am not sure its worth the extra calls.
Hopefully once we have live-migrate and cold-migrate/resize refactor
done we can add a new single API that better combines the different
approaches, which should remove the confusion.
Thinking about it, that should sove the issue about the different states too.
John
On 28 June 2013 09:18, guohliu <guoh...@linux.vnet.ibm.com> wrote:
On 06/27/2013 07:05 PM, John Garbutt wrote:
I haven't seen any plans to change this.
The way I see it, the states make most sense for resize, which is the
end-user facing operation.
Personally I see migrate as a more admin focused operation.
So to help simplify the code, I am OK with slightly confusing states
for those users.
The exception, I guess, is when users want to "re-balance" their
servers between availability zones.
With any luck, post refactor, it should be easier to re-visit this,
and perhaps add those extra states.
John
On 27 June 2013 09:14, guohliu <guoh...@linux.vnet.ibm.com> wrote:
Greetings,
I apologize if this question was already covered and I missed it, as we
know
the migrate and resize share the same code path as well as instance
status,
notification message etc in current code base, somehow it might confuse
the
user when he/she perform migrate but get VERIFY_RESIZE status, and as far
as
I know this wasn't to be targeted in migrate refactor work, I just
wondering
to know do you think this is a issue? how about to separate migrate and
resize with different instance status and notification message as well as
task status? any comments would be appreciated.
Best Regards
Guohliu
_______________________________________________
OpenStack-dev mailing list
OpenStack-dev@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
_______________________________________________
OpenStack-dev mailing list
OpenStack-dev@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Hi John,
Thanks for your comments, it makes sense to me, the resize is all good, we
might just need to slightly separate migrate status based on resize code
logic, but one exception is we might need to add confirm migrate and revert
migrate in admin action. thoughts?
Regards
Guohliu
_______________________________________________
OpenStack-dev mailing list
OpenStack-dev@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
_______________________________________________
OpenStack-dev mailing list
OpenStack-dev@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
I prefer that we don't need confirm and revert for migrate, because it
doesn't like resize, server location is not a transparent info compare
to flavor type, when migrate finish, that's it, thoughts? I would like
to propose a patch for detail discussion.
Regards
Guohliu
_______________________________________________
OpenStack-dev mailing list
OpenStack-dev@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
