So, env var iff it propagates through v1 code - I'...
# pexe
enough-analyst-54434
03/19/2020, 6:46 PMSo, env var iff it propagates through v1 code - I'm not sure it does - would be PEX_ROOT=/tmp/random
w
witty-crayon-22786
03/19/2020, 7:13 PMit won't, because the outer pex (in our case) strips it
witty-crayon-22786
03/19/2020, 7:14 PMcc @wide-energy-11069: ^
w
wide-energy-11069
03/19/2020, 7:15 PMoh let me see if pants allows to configure this
w
witty-crayon-22786
03/19/2020, 7:15 PM@wide-energy-11069: is this related to the 1.26.x upgrade?
witty-crayon-22786
03/19/2020, 7:15 PMbecause things are quite different there.
witty-crayon-22786
03/19/2020, 7:15 PMvs 1.25.x
w
wide-energy-11069
03/19/2020, 7:16 PMyes, we need to a way to force the resolve, as folks asked us to test their server configs.
w
witty-crayon-22786
03/19/2020, 7:17 PMk. then make sure you answer the question in the context of 1.26.x. i can help you use our dev branch internally.
w
wide-energy-11069
03/19/2020, 7:19 PMjust checked that PEX_ROOT doesn’t work in this case.
w
witty-crayon-22786
03/19/2020, 7:24 PMalso, the resolver/pex parallellism setting is likely to be relevant.
witty-crayon-22786
03/19/2020, 8:32 PM@enough-analyst-54434: ... i can't seem to find what i was sure i had seen. there is now a pex parallelism/"workers" setting, right?
witty-crayon-22786
03/19/2020, 8:36 PM"jobs"! .. argh, sorry.
witty-crayon-22786
03/19/2020, 8:36 PMpython-setup.resolver_jobs
e
enough-analyst-54434
03/19/2020, 9:22 PMSo @wide-energy-11069 is this a new requirement just for testing the upgrade? Or were you previously able to do this with Pants/Pex - force a re-resolve that now you cannot?
w
witty-crayon-22786
03/19/2020, 10:00 PMpossibly due to the parallelism, and possibly due to changes in pip (relative to pex's old resolver), we were seeing the servers hosting the wheels hitting their limits