Help me debug my daily Indigo web interface crash

Q1: Is it possible that difficulty opening an image (in this case from an wires ip camera) could cause this kind of error?

Q2: Is there a way to detect the WebServer is shutting down (short of scanning the log)? Thanks Jay

A1 - yes, certainly. You're still opening a "file", it's just not a disk file. So you need to close the pipe or whatever you're using to read the image.

A2 - no idea.

Q1: Is it possible that difficulty opening an image (in this case from an wires ip camera) could cause this kind of error?

Q2: Is there a way to detect the WebServer is shutting down (short of scanning the log)? Thanks Jay

A1 - yes, certainly. You're still opening a "file", it's just not a disk file. So you need to close the pipe or whatever you're using to read the image.

A2 - no idea.

Flying Diver i think we have the culprit. The refreshing image location url supplies an image that is temporarily unable to close a pipe/socket whatever, and the refreshing nature causes one of these per second.

Since the source cameras are rotated every 10 seconds, this might undermine any existing error checking.

Or something. Matt? Jay? Flying Diver says it’s YOUR fault! (Totally joking) ;P

[jay (support)]

The refreshing image location url supplies an image that is temporarily unable to close a pipe/socket whatever, and the refreshing nature causes one of these per second.

Sorry, I'm not following you. Can you expand on your theory to identify what "things" are "temporarily unable to close a pipe/socket"?

The refreshing image location url supplies an image that is temporarily unable to close a pipe/socket whatever, and the refreshing nature causes one of these per second.

Sorry, I'm not following you. Can you expand on your theory to identify what "things" are "temporarily unable to close a pipe/socket"?

If you’re not following, then I’m probably making no sense. I was running with Flying Diver’s response to my theory. Please disregard!

Follow up- limitting the use of my script that rotates through refreshing urls (linked to cameras that supply jpeg's) appears to have fixed the problem.

If you post the script maybe we can help figure out why it's causing the problem.

If you post the script maybe we can help figure out why it's causing the problem.

Thanks! I’ll post it when i get home tomorrow. I make little-to-no use of error catching, so hopefully that’s the problem.

FWIW- it’s actually three scripts that control rotating camera feeds in a control page “video” (updating url jpg).
One adds target url’s to a dict key of active sources.
One removes the url from active sources.
One rotates through the active sources every 10 seconds.

I suspect the problem lies in the third.