search

convexengineering / SPaircraft

Models for commercial aircraft design
http://spaircraft.readthedocs.org
25 stars 17 forks source link

Failing tests on one machine #61

Closed mayork closed 7 years ago

mayork commented 7 years ago

@bqpd any idea what is going on here? Not clear why this one machine is failing tests while everything else is passing

https://acdl.mit.edu/csi/job/gpkit_ResearchModel_d8_Push/buildnode=reynolds-ubuntu16,optimizer=mosek/444/testReport/junit/gpkit.tests.from_paths/TestFiles/test_SP_aircraftpy_mosek/

1ozturkbe commented 7 years ago

This is odd, because all of the tests are passing at the moment... but Ned let us know why this might be the case if you have ideas.

bqpd commented 7 years ago

looks like it's passing now?

1ozturkbe commented 7 years ago

...no? Latest commit (1f7ae93e) still failing tests. It's odd; the error is a NEAR_OPTIMAL and then an UNKNOWN, whereas the model runs fine in every other machine...

mayork commented 7 years ago

sometimes the machine passes and sometimes it fails. very odd.

bqpd commented 7 years ago

Going to run it a few times and see if the failure is only on reynolds-ubuntu16 or what.

bqpd commented 7 years ago
Trial Result
0 Failure on reynolds-ubuntu16
1 All passed
2 All passed
3 All passed
4 All passed
mayork commented 7 years ago

is something wrong with that machine?

bqpd commented 7 years ago

Maybe so? It seems to be either inconsistent, gone, or only happens on the first solve. (I'm still running more trials)

bqpd commented 7 years ago

Doesn't seem to be inconsistent...let me know if this happens when you do another push and we can see if it's a boundary condition or is just gone.

mayork commented 7 years ago

closing for now since we currently don't on Master..will reopen if observed again later