-
Notifications
You must be signed in to change notification settings - Fork 66
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
creating pngs in background mode leaks (much less) memory #2040
Comments
@aashish24 @doutriaux1 I created this follow-up issue if we want to dig more into this. |
@danlipsa I think this should be reviewed alongside #1424 as the problems may be linked. I'm just running my #1424 test script against |
#1424 will remain unresolved for now, I wasn't able to get a pull request in before the release. I'll try and get something together shortly so it can go into nightly builds. For the present, I'd recommend just using the "donotstoredisplay" argument. |
Hi Charles @charlesdoutriaux , Here is our repo for acme_diags: https://github.com/zshaheen/acme_diags/blob/master/docs/quick-guide-aims4.ipynb If you do below (which create a lot of pngs) Process gets slower and slower. No problem for cartopy backend... Also we have this libGL error when testing on aims4. |
@chengzhuzhang does your behavior look similar to #1424 (comment)? So you get a progressive slow down after each loop step? Oh and @doutriaux1 is the best github handle for Chucky (@charlesdoutriaux is his personal/non-work handle) |
@chengzhuzhang I'm fairly sure we're getting the libGL error b/c we aren't using vcs-nox. I'll try with that and we'll see what's good. It gets slow b/c of the libGL error, vcs in and of itself isn't causing that. @doutriaux1 No need to run it. If we're still getting errors, we'll let you know. |
@durack1 Thank you for the hint in poking Chucky :) |
This is a continuation of
#1397
While our VTK fix improved the leak considerably (from 500 MB to 6MB for 500 iterations) there is still a memory increase. It is possible that certain resources are still leaked.
More investigation is required to create a test that pinpoints the problem.
Here is the test program I used on linux:
png-memory.zip
The text was updated successfully, but these errors were encountered: