Closed edgetriggered closed 8 months ago
Thanks for your report!
1) We can swap out your laptop. That might be the easiest fix here
2) We'll think over potential fixes here. Maybe a config directive that extends timeouts
I've added a new config directive you can add to ~/.pyuscope. Sample entry that increases timeouts from 3 seconds to 30 seconds. Can you pull latest code, add this directive, and let me know if this solves your issue?
NOTE: another memory corruption issue was also identified and fixed.
{
"timeout_scalar": 10.0,
"labsmore_stitch": {
...
}
Is this possibly related to the memory leak? That should now be fixed: https://github.com/Labsmore/pyuscope/issues/375
Large scans are dying on loaner laptop (labtop3) at 75% point, seemingly caused by CPU frequency throttling having an effect on meeting a task deadline with three second timeout. Smaller test jobs with a few images to stitch run to completion.
dmesg is filled with messages like the following:
pyuscope.log