[PKG-Openstack-devel] Status?

[Thomas Goirand]

On 07/13/2017 10:12 PM, Ondrej Novy wrote:
> Hi,
> 
> 2017-07-13 17:26 GMT+02:00 Thomas Goirand <zigo at debian.org
> <mailto:zigo at debian.org>>:
> 
>     What your proposing is even worse than what I thought. Not only you wont
>     have a "build on each commit" type of thing (which I setup using Jenkins
>     and a receive hook), now you're even proposing to blindly upload without
>     even trying to build first!!! So no, "source-only" uploads isn't a
>     satisfying answer here, IMO.
> 
> 
> OMG Thomas, please be constructive! I DIDN'T wrote blindly upload
> without trying to build. Of course  I you need to build package locally,
> test it and then do source-only upload.

Then I fail to understand why you're talking about source-only upload
when I was talking about setting-up a CI. This makes no sense. Could you
explain how the 2 are related?

> Yes we will not have "build on each commit". But now we have "fail after
> every commit". What is better?

I was talking about setting-up a recieved hook type of CI, not using
what I've setup on OpenStack infra, since you wrote you don't want to
use it anymore. Now, you're complaining that the OpenStack infra is
failing (do I understand this bit correctly?), though why should I
attempt to fix the sources.list (which is a simple commit), if you're
not going to use it anyway?

> It's useless to compare your not working better solution with my. I'm
> trying to found "any working solution".

I'm sorry if you perceived it as comparing. That's now what I've been
trying to do.

>     On what hardware will this run? My last setup was with PXE booting a
>     Debian live system on which a Jenkins job was running a script using
>     ssh. The script was using openstack-deploy to setup an all-in-one node.
>     A VM would work, provided there's no MAC address filtering, and if
>     there's enough RAM, plus if you can "reset" it on each tempest job run.
> 
> we will see. I don't have answer for all questions now. But we need to
> start somewhere. tempest, CI, etc. is not must have. Many many package
> maintainers live without them.

I do not agree with you at all here. It is my point of view that running
tempest and validating packages is a must have for OpenStack. On each
and every release, I have seen failure to functional test OpenStack, and
running tempest was super helpful. Not doing any type of functional
testing is a huge regression, which would lead me to advise against
using OpenStack in Debian if someone asks.

Also, I don't think you can compare OpenStack with other packages. It's
quite unique in Debian to have such a large amount of packages for a
single software solution, and I don't know anything else that interact
with so many parts of the system (including, but not limited to: KVM,
Qemu, libvirt, rabbitmq-server, Maria/MySQL server, networking bridges,
etc.).

If not using an automatic deployment of OpenStack and tempest, how will
you make sure that the next release is really working? I see only a
single other way: following the install-guide (which may, at the same
time, provide you an opportunity to maintain it) and manually install
OpenStack, then manually run tempest. This could work, but would need a
lot more time than using the automation already in place.

>     I see no other way but to ask DSA for a VM here. The normal jenkins VMs
>     wont have enough RAM to run OpenStack, unfortunately.
> 
> yep, that can be one solution, but talk about it in future.

Probably Debconf would be a good place/time to discuss it with DSA.

> We have only few choices now
> 
>  1. try my solution
>  2. take care about packages yourself - you stated you can't/don't want
>  3. found someone which are going to do it your way - anyone?

I do not think that there's my way vs your way. I don't think anyone can
even remotely contest the advantage of functional testing. The only
issue is how, and if there's enough resources to do it.

> I don't want to write long flames with hates. I have way how I think it
> will work. You think it will not. We can try it, or do nothing. If we
> will do nothing, many packages will be AUTORMed if you are not going to
> take care of them.

The reason why I decide to stop doing the OpenStack packaging in Debian,
is because I cannot be fulltime on it, and I saw that there's not enough
contributor. Therefore, my thinking is that quality cannot be at the
level I think it should be. If you can't, or don't want to run tempest
and functional test packages, then I would advise you to not attempt to
continue the packaging of OpenStack in Debian as well. There's nothing
personal here, and nothing about your skills (you are a very skilled
person).

Cheers,

Thomas Goirand (zigo)