n7qnm
Junior Member
Posts: 80
|
Post by n7qnm on Nov 26, 2021 8:38:30 GMT -8
I'm playing with MQTT, trying to feed data from SkyWeather2 to weewx.
I have the basics working; mosquitto broker running on the "target", able to see messages using mosquitto_sub; but, I'm seeing some "interesting" behavior. I'm working on debug and isolating what's goin on; but wanted to run some things by the group.
First - I'm seeing messages from the FT020/AIO and the FT016 in addition to the "state" messages; and it looks like there are (at least) two connections being made to the broker. Was there some reason for that (rather than just the state dump)?
More importantly - after what seems to be a random amount of time, varying from immediately after startup to several hours, the MQTT messages stop, AND the "main" code becomes "deaf" to the RTL feed. In the debugging output, I see, "patting the dog", the SkyCam picture stuff, the state dump; but NO incoming data and NO MQTT messages.
AS I said, I'm still debugging - but would appreciate any insight....
Thanks!
Clay Jackson
|
|
n7qnm
Junior Member
Posts: 80
|
Post by n7qnm on Nov 26, 2021 12:17:57 GMT -8
Update - I turned off MQTT and have been watching the debug output. Data stopped about 10 minutes ago. I killed SkyWeather2 and started RTL_433 directly, no data! At the same time, I checked another Pi have have reading data, and it was seeing both sensors just fine (couldn't decode them, because it's mainline RTL, but it definitely saw the signals. Then I rebooted my SkyWeather Pi, and it's now receiving both sensors just fine.
|
|
|
Post by SDL on Nov 28, 2021 10:41:59 GMT -8
Look at the log file and see if you are seeing a bunch of rtl433 restart messages.
Rebooting daily is an option. Or even rebooting if the data stops.
BP
|
|
n7qnm
Junior Member
Posts: 80
|
Post by n7qnm on Dec 3, 2021 11:31:41 GMT -8
OK - I have the box inside now, and have been running "raw" rtl_433 for close to 24 hours. No crashes (I think that may have been a hardware issue, have to confirm and then will post); but, I do notice that the AIO reports are sometimes as much as an hour apart. How often does the AIO normally transmit?
I noticed a lot of "bit_offset" and "bit_buffer" lines on stdout, from switchdoclabs_SM23.c - I don't have one of those (yet), and that looks like debugging code anyway, so I commented out the "printf"s and recompiled. I'm still seeing "pulse_FSK_detect(): Maximum number of pulses reached!" fairly often and wondering if that's indicative of an issue somewhere. From discussions on the RTL boards, it looks like one possibility is interference, which is quite likely, I have at least 10 Accurite 433 devices.
Clay
|
|
|
Post by SDL on Dec 5, 2021 18:27:21 GMT -8
Interference (with 10 more devices) is absolutely a possibility. If I remember right it is about every minute from AIO.
BP
|
|