Forums

The goal is to turn a styled dataframe to a png that I could use later. To do this I use dataframe-image package that let's me do exactly that. I tested it before on my computer to check everything etc. and it worked without a problem.

Now I wanted to use scheduled task to make a png every day. Generating pngs using pyplot works great, but particularly with this package it doesn't work.

[0223/191701.822772:WARNING:gpu_process_host.cc(1296)] The GPU process has crashed 2 time(s)
[0223/191701.827303:ERROR:gpu_init.cc(426)] Passthrough is not supported, GL is disabled
[0223/191701.854410:WARNING:gpu_process_host.cc(1016)] Reinitialized the GPU process after a crash. The reported initialization time was 0 ms
[0223/191702.690438:ERROR:validation_errors.cc(115)] Invalid message: VALIDATION_ERROR_DESERIALIZATION_FAILED
[0223/191702.690839:ERROR:browser_child_process_host_impl.cc(747)] Terminating child process for bad message: Received bad user message: Validation failed for viz.mojom.CopyOutputResultSender.0  [VALIDATION_ERROR_DESERIALIZATION_FAILED]
[0223/191702.822221:WARNING:gpu_process_host.cc(1296)] The GPU process has crashed 3 time(s)
[0223/191702.822255:FATAL:gpu_data_manager_impl_private.cc(415)] GPU process isn't usable. Goodbye.

This is the error I've been getting, but lately another part showed up looking like this.

#23 0x55ff2d11a022 (/usr/lib/chromium/chromium+0x5b73021)
#24 0x7feb1406f9ba (/usr/lib/x86_64-linux-gnu/libevent-2.1.so.6.0.2+0x229b9)
#25 0x7feb14070537 event_base_loop

It's only part of the error. It's much too long to paste it all here. Does anyone know what to do? Is there a solution to that error or should I start looking for other package to turn df into a png and hopefully work here?

Our machines do not have a GPU, so if your code relies on a GPU, it will not work. It appears that you are using chromium at some point and for that, you will need to configure it to not try to use a GPU. We have example code here to show how to do that with selenium.

If anyone had the same problem, use df2img instead of dataframe-image.

Thanks for sharing!