Getting rid of Landfill?

[Jeff Fearn]

Not a direct response, but this touched on an idea that's been
flitting around my mind for a while, should BZ go DevOps?

One of the things that really annoys me about old school release
cycles is the large time between features being finished and them
being released. Every DevOps books is filled with the reasons on why
long lead times and huge payloads make software updates expensive and
fragile, so I won't cover those, but it'd be nice to not have to deal
with them.

The reason this email touches on it is because if you do go DevOps
then having a testable public instance on your trunk is extremely
useful and if Landfill is going to be rebuilt, then it'd be good to
take a potential switch to DevOps in to account when doing that.

AFAICT BMO have all the things in place to do DevOps with BZ, CI etc,
what's left to do is an agreement on the specifics of how to
commit/test/qe/release/etc.

I know a lot of work has gone in to 6, but if we go DevOps I'd like us
to consider reverting to 5 as the trunk and then pulling over feature
from 6 to 5 using the DevOps process to get the new stuff released.
This will serve as a reasonably robust test of the processes.

I'm sure some stuff will need to be reworked but a lot of it should be
fairly clean and we'd need to do some management of that list and
process. I'm pretty sure I can get some buy-in from work so I could
spend some work time pulling over content once the details have been
worked out.

Thoughts?

Cheers, Jeff.


On Thu, May 25, 2017 at 6:31 AM, Gervase Markham <gerv at mozilla.org> wrote:
> For a long time, the Bugzilla project has run Landfill
> <http://landfill.bugzilla.org/>, a testing and development server. It's
> had various roles over the years, but right now it just runs some
> installs of Bugzilla that people can use to try it out. It doesn't run
> tip installs because "Tip requires Perl 5.14 that's not readily
> available for CentOS 5." This has been the case for some time now, and
> doesn't look like getting easily fixed. There's been a sort of plan to
> upgrade it but no-one has found the time for months.
>
> In 2017, do we need this? Developers don't need it - if they need a
> Bugzilla, they just install one locally, in a VM of some sort if
> necessary, or use Docker, or whatever. Is it valuable enough to
> potential users of Bugzilla?
>
> The 5.0 install on Landfill seems to have quite a lot of use, but it
> seems to be mostly by people using it for their own bug tracking, or who
> are very confused.
>
> The reason to get rid of it is that it's barely maintained, and so a
> possible security risk, and worrying about it takes up developer and
> volunteer time that could better be spent elsewhere. If we did want a
> test server, perhaps a simple one running the latest stable version, why
> don't we just fire up a VM and install one on a modern OS, rather than
> trying to rebuild the current landfill box?
>
> Input on what we should do here is welcome :-)
>
> Gerv
>
> _______________________________________________
> dev-apps-bugzilla mailing list
> dev-apps-bugzilla at lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-apps-bugzilla
> -
> To view or change your list settings, click here:
> <https://lists.bugzilla.org/cgi-bin/mj_wwwusr?user=Jeff.Fearn@gmail.com>



-- 
The fundamental cause of the trouble is that in the modern world the
stupid are cocksure while the intelligent are full of doubt.
— Bertrand Russell, ‘The Triumph of Stupidity’, 1933/05/10, *Mortals
and Others: Bertrand Russells’ American Essays*, 1931–1935.