| 1 |
On 12/04/2014 04:08 PM, Rich Freeman wrote: |
| 2 |
> On Thu, Dec 4, 2014 at 2:36 PM, Alec Ten Harmsel |
| 3 |
> <alec@××××××××××××××.com> wrote: |
| 4 |
>> It'd be cool if Gentoo had some sort of automated workflow (with |
| 5 |
>> Jenkins, buildbot, whatever) like this: |
| 6 |
>> |
| 7 |
>> 1. Receive pull request |
| 8 |
>> 2. Detect changed ebuild |
| 9 |
>> 3. test ebuild with etest |
| 10 |
>> |
| 11 |
>> This will take a lot of work to set up, and depending on how my workload |
| 12 |
>> is next semester, I would love to help out as much as possible. |
| 13 |
>> |
| 14 |
> I've wondered the same thing, and maybe after sufficient bikeshedding |
| 15 |
> something like this would be a good GSoC project as well. Running |
| 16 |
> repoman is another possibility. |
| 17 |
|
| 18 |
Assuming I have time next semester, I'll look into it (and if I haven't |
| 19 |
mentioned anything by mid January, someone should bug me). I have |
| 20 |
limited hardware - a single 6-core/32GB box to do testing - but I should |
| 21 |
be able to do a lot of testing on small packages. |
| 22 |
|
| 23 |
> One trick is that a pull request could impact multiple packages |
| 24 |
> (eclass changes, dependency updates, etc). |
| 25 |
|
| 26 |
This would not be terribly difficult to detect, but insanely expensive |
| 27 |
resource-wise. The only thing that comes to mind immediately is ensuring |
| 28 |
that the software that detects changes associates a set of packages with |
| 29 |
an eclass in some config file, and triggers rebuilds for those packages. |
| 30 |
|
| 31 |
> In these cases running |
| 32 |
> repoman probably adds more value than full tests. Either way there |
| 33 |
> will potentially be a high cost to do the testing, which could create |
| 34 |
> a backlog if we queue things up (not queuing things creates other |
| 35 |
> problems). |
| 36 |
|
| 37 |
I sync weekly, and yeah, there'll need to be some reasonably serious |
| 38 |
hardware to do this. Jenkins looks like it supports adding workers that |
| 39 |
are firewalled by having the workers initiate connections; if this is |
| 40 |
the case (or if whatever we end up supports this), I'll gladly offer up |
| 41 |
my workstation to do this. |
| 42 |
|
| 43 |
> Repoman false-alarms is another potential issue with this. |
| 44 |
|
| 45 |
Repoman can be improved ;) Repoman is already run nightly, I just don't |
| 46 |
think Gentoo has the manpower to deal with that stuff. |
| 47 |
|
| 48 |
Alec |
Archives