image_processing crashes docker when reaching scan_interval (GPU docker)

[jaburges]

image_processing seems to send all entities at once to the docker container once scan_interval is reached. access via HA and curl confirms container no longer responds to queries

[jaburges]

i've raised an issue in the deepstack forum

workaround is to set a very very high scan_interval but knowing that in that amount of time you'll need to reboot deepstack docker.

Apparently other platforms can handle the simultaneous scan of multiple images, so thing this is a deepstack issue.

[robmarkcole]

Ok this is a deepstack issue. Really HA should be sending images for processing on update of the image, but a workaround is to stagger your scan_intervals so they dont coincide