Random lights turning on...solved

[mnsandler]

thanks for reporting back. glad you fixed it.

[warren@wcnjsmith.com]

SOLVED the problem. I was polling my 30+ devices at the rate of 1 second each - apparently just totally overloading the communications within the plugin and HS4. HS$ apparently became confused once in awhile and there the 'lets turn on group 254" message.

Since turning off polling - no odd issued with group 254 being activated. Response time is much better too.

Thank you and plugin working just great now.

[mnsandler]

SetIOMulti is what HS4 uses to control insteon devices. the plugin does not call that itself

something from hs (mobile app, alexa, hs event, etc) is calling the plugin

[warren@wcnjsmith.com]

Yes, indeed multiple times on reset and interface update.


Any ideas why this command would get transmitted from the plugin. No events triggered (only have 3 set up) , no groups set up. Only happened once yesterday and that was when the devices turned on at the random time.

What is SetIOMulti and TransmitGroupMsg ? When would these commands be sent and under what conditions.

Many thanks,
Warren



5/22/2022 09:52:40
Insteon
Info
TransmitGroupMsg called for device code: Group:254, Hub Group 254


5/22/2022 09:52:39
Insteon
Info
TransmitGroupMsg called for device code: Group:254, Hub Group 254


5/22/2022 09:52:39
Insteon
Info
Processing SetIOMulti for Insteon Group - Group Controller Cmd: 0


5/22/2022 09:52:38
Insteon
Info
Processing SetIOMulti for Insteon Group - Group Controller Cmd: 0

[mnsandler]

did you every factory reset the hub and run the plugin interface/update interface?

[warren@wcnjsmith.com]

Only 3 events set up - sunrise, sunset and an 11pm shut off. No other triggers. Simple set up. No groups either. Wondering if there are no groups set up that perhaps after a period of time this happens - especially since its Group 254

I have turned off polling just to see if that was contributing.

This only happens randomly - every few days

[mnsandler]

that looks like the insteon global group. are you triggering that somewhere

To post an HS log if you have issues or get errors (Do not copy and post the log, its unreadable)
- Goto to the hs tools/log menu, Export the log of the transactions, download it and attach it to your post

[warren@wcnjsmith.com]

I did see this message in the log at the time of the Wireshark captured message. Only once did I see this message in today's log


5/22/2022 09:52:40
Insteon
Info
TransmitGroupMsg called for device code: Group:254, Hub Group 254


5/22/2022 09:52:39
Insteon
Info
TransmitGroupMsg called for device code: Group:254, Hub Group 254


5/22/2022 09:52:39
Insteon
Info
Processing SetIOMulti for Insteon Group - Group Controller Cmd: 0


5/22/2022 09:52:38
Insteon
Info
Processing SetIOMulti for Insteon Group - Group Controller Cmd: 0

[mnsandler]

Please correlate the wireshark time with the hs log.

the only commands the plugins sends are status polling commands and those are easily disabled/.

the plugin does not send commands by itself. something must be invoking it.

or some other application/system talking to the insteon hub

[warren@wcnjsmith.com]

I manually sent the command that I captured in wireshark - GET /3?02620000FECA1100=I=3

Yes, indeed the same 'random' lights came on. Seems that indeed The plugin is very occasionally sending this packet - once every few days it occurs .





Frame 2101881: 186 bytes on wire (1488 bits), 186 bytes captured (1488 bits) on interface \Device\NPF_{BE20422A-04FA-45DE-8465-5839DB188653}, id 0
Ethernet II, Src: Microsof_00:be:01 (00:15:5d:00:be:01), Dst: Smartlab_4a:8d:bd (00:0e:f3:4a:8d:bd)
Internet Protocol Version 4, Src: 10.0.0.211, Dst: 10.0.0.123
Transmission Control Protocol, Src Port: 65243, Dst Port: 25105, Seq: 1, Ack: 1, Len: 132
Source Port: 65243
Destination Port: 25105
[Stream index: 147781]
[Conversation completeness: Complete, WITH_DATA (31)]
[TCP Segment Len: 132]
Sequence Number: 1 (relative sequence number)
Sequence Number (raw): 2281471029
[Next Sequence Number: 133 (relative sequence number)]
Acknowledgment Number: 1 (relative ack number)
Acknowledgment number (raw): 1228355754
0101 .... = Header Length: 20 bytes (5)
Flags: 0x018 (PSH, ACK)
Window: 65392
[Calculated window size: 65392]
[Window size scaling factor: -2 (no window scaling used)]
Checksum: 0x15ec [unverified]
[Checksum Status: Unverified]
Urgent Pointer: 0
[Timestamps]
[SEQ/ACK analysis]
TCP payload (132 bytes)
Hypertext Transfer Protocol
GET /3?02620000FECA0000=I=3 HTTP/1.1\r\n
[Expert Info (Chat/Sequence): GET /3?02620000FECA0000=I=3 HTTP/1.1\r\n]
[GET /3?02620000FECA0000=I=3 HTTP/1.1\r\n]
[Severity level: Chat]
[Group: Sequence]
Request Method: GET
Request URI: /3?02620000FECA0000=I=3
Request Version: HTTP/1.1
User-Agent: Insteon\r\n
Authorization: Basic RG9yYW41ODM6QlZ6UUlvRVg=\r\n
Host: 10.0.0.123:25105\r\n
\r\n
[Full request URI: http://10.0.0.123:25105/3?02620000FECA0000=I=3]
[HTTP request 1/1]
[Response in frame: 2101884]

[warren@wcnjsmith.com]

Half dozen lights randomly came on this morning. Fortunately had a wireshark trace running. I saw a message that was inconsistent with the usual get status messages. Having a bit of trouble decoding it however


Usual status messages - 3?02642b8dd320a1900=i=3 - I assume the 19 and 00 are related to the command to get status of the specific device

Here is one that came across about the time of the random lights turning on - 3?02620000FECA1100 I believe the 11 00 is the command to turn on a device however 00 would indicate zero brightness ?
No activity was happening on any of the HS events at this time

Or perhaps I am barking up the wrong tree and decoding the messages incorrectly.

I have however reset the devices, even fully deleted them then on HS and the plugin and reenrolled them per the advice in prior messages. I do have a new hub on order (well new used) and will try that as well.

After this I now have turn on debug logging to see if we can see anything more.

Many thanks

Warren

[mnsandler]

from my post # 6
have you tried a factory reset on the device? Then just run the 'program device for homeseer' to relink it. do no delete the device from the plugin.

[claude]

Just so you feel better, I have one switch that turns on sporadically. I can't prove it, but it's probably circuit noise.

As Mark indicated, running detailed logging will eliminate Insteon protocol interventions. Running WireShark at this point is not beneficial IMO (it can be, but not for your situation).

Shots in the dark:

• Are the 6 devices all on the same electrical circuit?
• Have you tried resetting those 6 devices and reprogramming them? I have had one situation with a rogue device. Everything looked fine from an Insteon point of view. In desperation, did a hardware reset of the switch, problem gone. Now, the same situation with 6 devices would be unusual, but who knows what happens during brown outs
• Run a 'Display interface links'

Good luck

[mnsandler]

make sure to run the plugin interface/update interface after a factory reset

you could factory reset those devices, and then run the program device for homeseer again. you do no need to delete them from the plugin

enabled detailed logging on the plugin settings page and see if that captures anything

also check any Events you created

[warren@wcnjsmith.com]

still having the issue of random devices turning on. Yesterday, I installed the latest insteon version. Upon restart random lights turned on. However nothing in the log indicated that these lights were turned on by the plug in. I then stopped the HS4 service, restarted, but could not reproduce the problem. Hub has been factory reset multiple times. Runnign latest version of HS4 and Insteon.

ITs the same devices that get turned on each time. They are dome that are the DIN rail dimmers, some are just dimmer switches. I have cleaned up any dead links so all looks clean on the devices.

I was thinking about running wireshark to look at the tcp packets between HS4 and the hub, however since this happens so randomly it may be a futile logging effort.

Here is a wireshark capture running with conversation between HS4 and the hub

Thanks to all

Warren