HTTPCMD Timeout

[JeffS]

What are the chances this timeout could be editable?

I have some busier PLCs that seem to struggle doing some of the TLS stuff within the 30 second timeframe.  Typical scan times aren't crazy but average 11-12ms which based on wireshark data has the TLS response computation right at 30 seconds.  If I manually suspend programs to lower my scantime then the HTTPS completes within the 30 second timeout.

[BobO]

What are the chances this timeout could be editable?

I have some busier PLCs that seem to struggle doing some of the TLS stuff within the 30 second timeframe.  Typical scan times aren't crazy but average 11-12ms which based on wireshark data has the TLS response computation right at 30 seconds.  If I manually suspend programs to lower my scantime then the HTTPS completes within the 30 second timeout.

I'm sure it could be, it's hard coded now. At the least we could make it accessible via DEVREAD/DEVWRITE.

A big concern is how tolerant the server is when the TLS handshake goes long. Many servers drop the connection at 15 seconds.

[JeffS]

In this particular case the server is ours, so we can allow for a greater timeout.  I dug through my code to see if I could possibly suspend some programs but nearly all make sure of rising/falling edge contacts that would be missed while the program is suspended.  So at least for our situation I would really love a way to just extend the timeout so the PLC has more time to compute the response.