Shutdown subscription after exactly 30 minutes

[agners]

Instead of relying on the progressively growing and random resubscribe interval simply check the time since the first re-subscription attempt.

The resubscribe interval is chosen at random, and even in the terminal state, a resubscription potentially happens below 30 minutes (in the final progression the resubscription time is chosen at random between 27 and 92 minutes). This means that in theory we shutdown the subscription much later than actually intended.

[agners]

It seems that the callback provided time is "made up", and doesn't reflect the actual resubscription time, e.g.:

2024-09-04 10:20:59.046 (Dummy-2) CHIP_PROGRESS [chip.native.DMG] Will try to resubscribe to 01:0000000000000001 at retry index 5 after 29988ms due to error src/protocols/secure_channel/CASESession.cpp:560: CHIP Error 0x00000032: Timeout
2024-09-04 10:20:59.047 (MainThread) INFO [matter_server.server.device_controller] <Node:1> Previous subscription failed with Error: 50, re-subscribing in 21192 ms...
...
2024-09-04 10:21:29.035 (Dummy-2) CHIP_PROGRESS [chip.native.DMG] OnResubscribeTimerCallback: ForceCASE = 1

(the actual resubscribe time was after ~29.9s, so the value provided by the SDK, not the Python callback)

This change just removes the Python callback provided time, it is useless currently.