10:49 a.m.
On 10/21/2013 04:11 PM, Diogo Vieira wrote:
I have installed Openstack Grizzly (RDO distribution) and now want to
upgrade to the most recent version (Havana). Is there any information somewhere on how to
proceed?
dansmith or larsk may be able to assist. Moving this thread to rdo-list
though, since it is RDO related
Perry
11:03 a.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
Hi Diogo,
dansmith or larsk may be able to assist. Moving this thread to
rdo-list
though, since it is RDO related
I don't think that there are any targeted docs up anywhere yet. However,
I've been testing it myself to try to make sure we know what the best
plan of action is. I have a tool up that I use for testing this, and the
scenario files are fairly readable. This is the easiest and most
straightforward way to go:
https://github.com/kk7ds/test_rdo_upgrade/blob/master/scenarios/grizzly2h...
That has stuff in it for testing the before/after cases which you
probably don't want to do, but the steps are basically the same.
Shutdown all the services, upgrade the packages on all the nodes, then
run the database upgrades, and then start everything back up. Note that
the scenario above includes a re-run of packstack, but this is not
really necessary.
To do the database upgrades, each service has a "manage" tool that does
low-level database management, so for nova it is:
nova-manage db sync
This tool only needs to be run once per service and the order of
services doesn't matter.
Make sense?
I will try to get my notes up on a public wiki page somewhere so that
this is available someplace other than just this email thread.
--Dan
11:19 a.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
Thank you! Yes, it makes sense, I'll give it a try.
Diogo Vieira
On Oct 21, 2013, at 5:03 PM, Dan Smith <dansmith(a)redhat.com> wrote:
Hi Diogo,
> dansmith or larsk may be able to assist. Moving this thread to rdo-list
> though, since it is RDO related
I don't think that there are any targeted docs up anywhere yet. However,
I've been testing it myself to try to make sure we know what the best
plan of action is. I have a tool up that I use for testing this, and the
scenario files are fairly readable. This is the easiest and most
straightforward way to go:
https://github.com/kk7ds/test_rdo_upgrade/blob/master/scenarios/grizzly2h...
That has stuff in it for testing the before/after cases which you
probably don't want to do, but the steps are basically the same.
Shutdown all the services, upgrade the packages on all the nodes, then
run the database upgrades, and then start everything back up. Note that
the scenario above includes a re-run of packstack, but this is not
really necessary.
To do the database upgrades, each service has a "manage" tool that does
low-level database management, so for nova it is:
nova-manage db sync
This tool only needs to be run once per service and the order of
services doesn't matter.
Make sense?
I will try to get my notes up on a public wiki page somewhere so that
this is available someplace other than just this email thread.
--Dan
12:14 p.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
Dan,
Thanks for this effort. I'd fully support the public wiki page as it allows for
explanations and incremental improvements along with the script.
Have you had a look at what it would take to do a rolling upgrade rather than down all the
services ?
Tim
-----Original Message-----
From: rdo-list-bounces(a)redhat.com [mailto:rdo-list-bounces@redhat.com] On Behalf Of Diogo
Vieira
Sent: 21 October 2013 18:19
To: Dan Smith
Cc: Perry Myers; rdo-list
Subject: Re: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
Thank you! Yes, it makes sense, I'll give it a try.
Diogo Vieira
On Oct 21, 2013, at 5:03 PM, Dan Smith <dansmith(a)redhat.com> wrote:
> Hi Diogo,
>
>> dansmith or larsk may be able to assist. Moving this thread to
>> rdo-list though, since it is RDO related
>
> I don't think that there are any targeted docs up anywhere yet.
> However, I've been testing it myself to try to make sure we know what
> the best plan of action is. I have a tool up that I use for testing
> this, and the scenario files are fairly readable. This is the easiest
> and most straightforward way to go:
>
> https://github.com/kk7ds/test_rdo_upgrade/blob/master/scenarios/grizzl
> y2havana.sh
>
> That has stuff in it for testing the before/after cases which you
> probably don't want to do, but the steps are basically the same.
> Shutdown all the services, upgrade the packages on all the nodes, then
> run the database upgrades, and then start everything back up. Note
> that the scenario above includes a re-run of packstack, but this is
> not really necessary.
>
> To do the database upgrades, each service has a "manage" tool that
> does low-level database management, so for nova it is:
>
> nova-manage db sync
>
> This tool only needs to be run once per service and the order of
> services doesn't matter.
>
> Make sense?
>
> I will try to get my notes up on a public wiki page somewhere so that
> this is available someplace other than just this email thread.
>
> --Dan
_______________________________________________
Rdo-list mailing list
Rdo-list(a)redhat.com
https://www.redhat.com/mailman/listinfo/rdo-list
12:21 p.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
On Mon, Oct 21, 2013 at 05:14:22PM +0000, Tim Bell wrote:
Have you had a look at what it would take to do a rolling upgrade
rather than down all the services ?
Dan's other scenario does a service-by-service upgrade:
https://github.com/kk7ds/test_rdo_upgrade/blob/master/scenarios/grizzly2h...
In addition to the steps shown in that script, I would suggest running
a "yum update" after everything else has completed because that
script, as currently written, selectively updates only packages
specific to the service being upgraded (and so might miss updates to
basic system tools, etc).
-- Lars
--
Lars Kellogg-Stedman <lars(a)redhat.com>
1:23 p.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
Thanks for this effort. I'd fully support the public wiki page as
it
allows for explanations and incremental improvements along with the
script.
Yeah, I definitely need to get that up :)
Have you had a look at what it would take to do a rolling upgrade
rather than down all the services ?
Yeah. So, Lars just wrote back about the service-by-service approach,
which also works well and provides smaller downtime gaps and fewer
"committed" components during the upgrade if something goes wrong.
Further, it also lets you upgrade everything except nova, and then
deploy a new havana nova cluster that shares resources with the
still-on-grizzly deployment. With that, you can slowly migrate your
instances to the new nova deployment, growing that deployment as nodes
become vacated from the old one, eventually deprecating the old one
completely. That offers tenant, credential, image, and volume continuity
across the upgrade, as well as a much larger timeframe to migrate
everything, other than just "nova is down, move everything!!" :)
To me, "rolling upgrades" means something else, which I think is
probably what you're referring to. My upstream work right now is
targeted at making sure that we can upgrade nova components individually
without needing the parallel nova deployment above. We're getting ever
closer to this being a reality, and I _hope_ we will start gating on
"not breaking live upgrades" in the Icehouse or J timeframes.
--Dan
1:28 p.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
Any idea how cells would fit into this ?
Trying to work out the upgrade order for the Nova API cells vs the compute cells is one of
our open questions....
Tim
-----Original Message-----
From: Dan Smith [mailto:dansmith@redhat.com]
Sent: 21 October 2013 20:23
To: Tim Bell; Diogo Vieira
Cc: Perry Myers; rdo-list
Subject: Re: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
> Thanks for this effort. I'd fully support the public wiki page as it
> allows for explanations and incremental improvements along with the
> script.
Yeah, I definitely need to get that up :)
> Have you had a look at what it would take to do a rolling upgrade
> rather than down all the services ?
Yeah. So, Lars just wrote back about the service-by-service approach, which also works
well and provides smaller downtime gaps and
fewer "committed" components during the upgrade if something goes wrong.
Further, it also lets you upgrade everything except nova, and then deploy a new havana
nova cluster that shares resources with the still-
on-grizzly deployment. With that, you can slowly migrate your instances to the new nova
deployment, growing that deployment as nodes
become vacated from the old one, eventually deprecating the old one completely. That
offers tenant, credential, image, and volume
continuity across the upgrade, as well as a much larger timeframe to migrate everything,
other than just "nova is down, move
everything!!" :)
To me, "rolling upgrades" means something else, which I think is probably what
you're referring to. My upstream work right now is
targeted at making sure that we can upgrade nova components individually without needing
the parallel nova deployment above. We're
getting ever closer to this being a reality, and I _hope_ we will start gating on
"not breaking live upgrades" in the Icehouse or J
timeframes.
--Dan
1:44 p.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
Any idea how cells would fit into this ?
I don't, no. I would love to get to the point where Cells helps the
upgrade story, by supporting upgrade of individual cells, or adding a
new cell of Havana to make the situation I described easier to deploy.
However, I don't think enough people outside the CD camp are using, and
specifically, *contributing* to cells to make this a reality.
Trying to work out the upgrade order for the Nova API cells vs the
compute cells is one of our open questions....
Yeah, I don't think I have anything to offer for this at the moment. The
objects work we're doing upstream to make this work for nova itself will
definitely help make this a reality, but I think the relatively small
number of people deploying (and again, contributing to) cells means this
doesn't get a lot of attention :(
--Dan
5:29 a.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
Hi,
On 10/21/2013 08:44 PM, Dan Smith wrote:
> Trying to work out the upgrade order for the Nova API cells vs
the
> compute cells is one of our open questions....
Yeah, I don't think I have anything to offer for this at the moment. The
objects work we're doing upstream to make this work for nova itself will
definitely help make this a reality, but I think the relatively small
number of people deploying (and again, contributing to) cells means this
doesn't get a lot of attention :(
Tim, if you, Jan or anyone else has suggestions for improving
http://openstack.redhat.com/Upgrading_RDO I would be very happy to see them.
Do you have in-place rolling upgrades working in your set-up?
Thanks,
Dave.
--
Dave Neary - Community Action and Impact
Open Source and Standards, Red Hat - http://community.redhat.com
Ph: +33 9 50 71 55 62 / Cell: +33 6 77 01 92 13
10:37 a.m.
New subject: [Rdo-list] [rhos-list] Openstack Grizzly upgrade to Havana
On 10/21/2013 07:23 PM, Dan Smith wrote:
> Thanks for this effort. I'd fully support the public wiki
page as it
> allows for explanations and incremental improvements along with the
> script.
Yeah, I definitely need to get that up :)
This seems to be in place now:
http://openstack.redhat.com/Upgrading_RDO
thanks,
Pádraig.
4047
days inactive
4055
days old
9 comments
7 participants
participants (7)
-
Dan Smith -
Dave Neary -
Diogo Vieira -
Lars Kellogg-Stedman -
Perry Myers -
Pádraig Brady -
Tim Bell