Issue 19886: Better estimated memory requirements for bigmem tests (original) (raw)

This issue has been migrated to GitHub: https://github.com/python/cpython/issues/64085

classification

process

Created on 2013-12-04 19:30 by serhiy.storchaka, last changed 2022-04-11 14:57 by admin. This issue is now closed.

Messages (5)
msg205236 - (view)	Author: Serhiy Storchaka (serhiy.storchaka) *	Date: 2013-12-04 19:30
Some bigmem tests actually consume much more memory than declare. This can cause swapping and too long time of test run. Here is a patch which improves estimates.
msg205238 - (view)	Author: Antoine Pitrou (pitrou) *	Date: 2013-12-04 19:32
Why did you compute those estimates? When coming up with factors such as 3.3 or 3.57 (!), it would be nice to add a comment to explain the reasoning. (otherwise, thanks a lot for doing this)
msg205239 - (view)	Author: Serhiy Storchaka (serhiy.storchaka) *	Date: 2013-12-04 19:42
These are experimental values. I hacked tests and ran them with less sizes (using `ulimit -v` to hard limit memory size). And then I had approximated measured values. On Windows values can be larger (due different memory management).
msg207852 - (view)	Author: Roundup Robot (python-dev)	Date: 2014-01-10 11:41
New changeset aee672ba9c38 by Serhiy Storchaka in branch '2.7': Issue #19886: Use better estimated memory requirements for bigmem tests. http://hg.python.org/cpython/rev/aee672ba9c38 New changeset 2ca26065fb00 by Serhiy Storchaka in branch '3.3': Issue #19886: Use better estimated memory requirements for bigmem tests. http://hg.python.org/cpython/rev/2ca26065fb00 New changeset fff1455e1147 by Serhiy Storchaka in branch 'default': Issue #19886: Use better estimated memory requirements for bigmem tests. http://hg.python.org/cpython/rev/fff1455e1147
msg207855 - (view)	Author: Serhiy Storchaka (serhiy.storchaka) *	Date: 2014-01-10 11:50
I never decreased requirements, so these changes shouldn't cause regressions.