These are chat archives for astropy/astropy

15th
Jun 2017
matrixbot
@matrixbot
Jun 15 2017 19:05
rohit on Freenode hey guyz...
matrixbot
@matrixbot
Jun 15 2017 20:59
saimn on Freenode bsipocz: hey, you here ?
saimn on Freenode how does it work for the travis queue, just to be sure I understand correctly:
bsipocz hey
saimn on Freenode only one build is running, both for master (merges) and pr ?
saimn on Freenode it's taking >3 hours per build now !
bsipocz we have 5 threads running simultaniously (that will go up to 15 tomorrow, I hope at least).
bsipocz and everything that is a PR or a branch push under the astropy organization goes into the same queue
saimn on Freenode ah ok, so 5 jobs ?
bsipocz yes
bsipocz however I would suggest to cancel out the OSX ones, as travis is overloaded there, so the jobs are waiting an extra ~30-60 mins in another queue when on OSX
matrixbot
@matrixbot
Jun 15 2017 21:04
saimn on Freenode indeed
bsipocz so it's rather bad atm, we have ~4 hours of backlog
saimn on Freenode I was wondering why the latest build took more than 3 hours
saimn on Freenode it was more around 2 hours a few days ago
bsipocz if you have something that needs to be tested and adjusted, I suggest to enable travis on your fork so you don't need to wait the 4 hours
bsipocz yes, I saw that, too, actually not sure whether that's just a one off travis issue that we get slower stuff in the cloud, or indeed more tests are added that slow things done. Sadly the benchmarks are not yet running
matrixbot
@matrixbot
Jun 15 2017 21:12
bsipocz hmm
bsipocz you must be right, something is going rather bad
bsipocz @mirca are you here?
saimn on Freenode 4 tests that takes 30sec, times the number of jobs ...
bsipocz I'm afraid astropy/stats/tests/test_spatial.py::test_ripley_large_density[points0-0-1] from your K fucntions are burning too much CI time
bsipocz Thanks saim, happy to see you got to the same conclusion, but faster and nicer
matrixbot
@matrixbot
Jun 15 2017 21:19
bsipocz saimn: do you plan to open an issue for it or shall I?
Zé Vinícius
@mirca
Jun 15 2017 21:20
@bsipocz I’m here!
matrixbot
@matrixbot
Jun 15 2017 21:20
bsipocz great!
Zé Vinícius
@mirca
Jun 15 2017 21:20
No problem, do you want me to reduce the size of the input in that test?
matrixbot
@matrixbot
Jun 15 2017 21:21
bsipocz your test_spatial.py needs to be sorted out, it takes 4.5 mins on my computer to run, and as saimn points out almost an extra our in total for travis
bsipocz whatever that works and still sensible
Zé Vinícius
@mirca
Jun 15 2017 21:21
Sure!
Give me a minute
matrixbot
@matrixbot
Jun 15 2017 21:24
saimn on Freenode I can open the issue, is it still needed ?
bsipocz nope, we got mirca's attention here :)
saimn on Freenode ok great!
saimn on Freenode btw it could be interesting to have the slowest tests (50?) on one of the travis jobs
matrixbot
@matrixbot
Jun 15 2017 21:30
bsipocz how much is the overhead to get that stat? If it doesn't add much, I think it's an excellent idea
bsipocz I mean that something must have gone wrong in https://travis-ci.org/astropy/astropy/builds/243197060
saimn on Freenode nice graph :)
bsipocz also probably the unit support in modelling also added some extra time
saimn on Freenode yeah probably
saimn on Freenode I don't know exactly the overhead to get the tests timings, but it would also allow to know the real timing for each slow test
saimn on Freenode depending on their cpu I mean
saimn on Freenode for me the slowest takes 30sec per test, it's probably more on travis ;)
matrixbot
@matrixbot
Jun 15 2017 21:35
saimn on Freenode I will test on my fork
bsipocz Probably it takes longer on travis, that file containing the four slowest was much longer on mine:
real 4m23.816s user 4m18.258s sys 0m2.739s
Zé Vinícius
@mirca
Jun 15 2017 21:40
@bsipocz done :) astropy/astropy#6224
matrixbot
@matrixbot
Jun 15 2017 21:41
bsipocz perfect, thanks!
saimn on Freenode thanks @mirca !
Zé Vinícius
@mirca
Jun 15 2017 21:44
=D
matrixbot
@matrixbot
Jun 15 2017 22:48
saimn on Freenode here that ~120s per test
saimn on Freenode so 8 min per job
saimn on Freenode have to go to bed, so I will push a pr with --durations later
bsipocz that explains the +1 hour
saimn on Freenode also there is a weird interaction between --durations and --mpl : https://travis-ci.org/saimn/astropy/jobs/243453988#L625
bsipocz Thanks. That PR is not critical for the freeze, so no rush.
saimn on Freenode so I added the option in a build without --mpl
saimn on Freenode yep
bsipocz sounds good.
saimn on Freenode bye
bsipocz good night
matrixbot
@matrixbot
Jun 15 2017 22:53
bsipocz @mirca - you know that you can opt in to run travis on pushes, so it will run when you add a commit to any of your branches?
bsipocz (no need to open a PR against your fork)
Zé Vinícius
@mirca
Jun 15 2017 22:56
yeah ;)