On Sun, 23 Jun 2013 20:40:17 -0400 Jeffrey Hutzelman <[email protected]> wrote:
> On Sat, 2013-06-22 at 15:12 -0400, Jason Edgecombe wrote: > > > > For platforms with multiple slaves, we could have two builders, one with > > all the slaves, and one with only one slave. The single-slave builder > > could be assigned to the master branch and the multi-slave builder could > > be assigned to the other branches. > > I guess I don't see an advantage to this. The branches which only get > pullups once in a while get all the cycles, while the branch where the > real work happens gets to be starved? > > The whole reason this came up is because I noticed it was taking a long > time for the builtbot to get to changes I submit, even when things are > relatively quiet, and I wanted to throw more cycles into the pool to > make that get better. In addition to more cycles per seconds in the pool of build slaves, we could make the buildbot master configuration a bit smarter about about scheduling builds from gerrit. Changes restricted to src/WINNT could trigger only the windows builders, and changes restricted to src/afs would trigger only unix-y builders. I would be happy to help with these configuration changes. Secondly, I think it might be interesting to setup a new scheduler to run the new tap tests periodically (perhaps nightly) on a set of new build slaves. Thanks, Mike -- Michael Meffie <[email protected]> _______________________________________________ OpenAFS-devel mailing list [email protected] https://lists.openafs.org/mailman/listinfo/openafs-devel
