[petsc-dev] Improving nightly builds for maint

[Lisandro Dalcin]

On 21 October 2017 at 19:32, Jed Brown <jed at jedbrown.org> wrote:
> Satish Balay <balay at mcs.anl.gov> writes:
>
> Depends if you're using the machine for other things.  I think having
> that long-running job would tend to oversubscribe MPI and slow down
> throughput.  In any case, we probably shouldn't run this 3D convergence
> study under any testing system, even without Valgrind.
>

Indeed. I'm also in charge of maintaining the Travis-CI builds. Given
that we do not pay them, I think it is abusive to overload their build
workers.

PS: As it is not my intention to impose my requirements on anyone
else, I'm going to merge the branch github/petsc/ci/travis-ci
disabling the test runs on Travis-CI. The OS build workers are heavily
overloaded, sometimes we have to wait hours for builds to finish, and
the test runs, because of the way they are done (`make` does fails
with non-zero exit status), are a bit useless. Let's keep Travis-CI as
a quick way to get fast feedback on commits that would make the build
fail, a no more. Anyway, we all depend on merging into next and
checking the nightly builds.


-- 
Lisandro Dalcin
============
Research Scientist
Computer, Electrical and Mathematical Sciences & Engineering (CEMSE)
Extreme Computing Research Center (ECRC)
King Abdullah University of Science and Technology (KAUST)
http://ecrc.kaust.edu.sa/

4700 King Abdullah University of Science and Technology
al-Khawarizmi Bldg (Bldg 1), Office # 0109
Thuwal 23955-6900, Kingdom of Saudi Arabia
http://www.kaust.edu.sa

Office Phone: +966 12 808-0459