Closed dummys closed 3 years ago
Can you reproduce this problem by:
If I stop and restart, I don't get the duplicate error, because I guess my pan_id is not the same as all my deployed device, right ?
Another problem if I want to follow the best practice and use the full serial device name:
Zigbee2MQTT:error 2021-06-12 21:17:02: Error: Error while opening serialport 'Error: Error: No such file or directory, cannot open /dev/serial/by-id/usb-Silicon_Labs_slae.sh_cc2652rb_stick_-_slaesh_s_iot_stuff_00_12_4B_00_21_4F_24_0F-if00-port0
and the name of the serial device is correct, if we look inside the terminal:
@dummys when doing this the pan id should be the same, can you check?
no because now I changed the pan id. and yes all my device are back but I had to repair ALL of them...
So what if you don't change your pan id? (procedure of https://github.com/Koenkk/zigbee2mqtt/issues/7761#issuecomment-860754446)
So what if you don't change your pan id? (procedure of #7761 (comment))
not possible, If I don't change the service didnt start
It's probably the same as https://github.com/Koenkk/zigbee2mqtt/issues/6302 , try reflashing the firmware with a different computer.
already tried, same problem. I re-paired everything :( Lets close this ticket. anyway, thanks for help
@Koenkk , sorry to re-open but thats crazy man. I have just restarted my home assistant Odroid device, without any UPDATE/change to z2m config or whatever, and got AGAIN The F***** pan id error. WTF WITH THIS DEVICE SRSLY ? Last time I setuped new PAN_ID, repaired all. I really don't want to do that each time I reboot my HA device to be honest. WHere is the bug ? I bought the SLAE because in your documentation it stated that it is the best device, but it is a nightmare to repair all device. I also tried your stuff with removing external antenna, using foil on the dongle, nothing it wont start at all ...
EDIT: Tried reflashing coordinator firmware just in case, SAME problem, and I cross check with terminal there is the coordinator_backup.json in Z2M directory.
EDIT2: When I shutdown all the electricity in house and keep under UPS the odroid, the zigbee2mqtt start but the database is empty, even when i first copy my database.db from the backup I made after all the first crap I get. Why the device is deleting the database and recreating one ?
What I tried lastly: Stop z2m reprogram firmware of dongle using python script recommended Here is the log of the reprogram:
-> $ python3 cc2538-bsl.py -p /dev/ttyUSB0 -evw CC2652RB_coordinator_20210120.hex
Opening port /dev/ttyUSB0, baud 500000
Reading data from CC2652RB_coordinator_20210120.hex
Your firmware looks like an Intel Hex file
Connecting to target...
CC1350 PG2.0 (7x7mm): 352KB Flash, 20KB SRAM, CCFG.BL_CONFIG at 0x00057FD8
Primary IEEE Address: 00:12:4B:00:21:4F:24:0F
Performing mass erase
Erasing all main bank flash sectors
Erase done
Writing 360448 bytes starting at address 0x00000000
Write 104 bytes at 0x00057F988
Write done
Verifying by comparing CRC32 calculations.
Verified (match: 0x927c0d4a)
restore snapshot put back dongle restart z2m -> No device at all... When I look with SSH, database.db is empty database.db.backup is full ... I would like to know how I can debug this. I have to leave home for week and my wife will be alone ... PS: I didnt do any update/upgrade since then... For me there is definitely a bug inside the SLAE firmware, because I do exactly what you asked for and no device joined.
First make sure you are on at least 1.19.0 (big changes to backup mechanism there).
Next reflash your stick and change the pan_id
and start (new network will be created).
Now connect pair one router device (e.g. a bulb, not a battery powered one). Next restart z2m, after this is the bulb still in the database?
First make sure you are on at least 1.19.0 (big changes to backup mechanism there).
To be honest I don't see why I would upgrade as the existing one must work. I don't want to change stuff before fixing, or understanding why I get this. To be honest I lose so much time with this, all my house is not working everytime update/or restart of HA/Z2m. I changed from CC to slae back in the days because in your page it was the recommenced one, I bought two of it and I don't get the point of this pan_id error. When a coordinator create a zigbee network, and there is router, I don't get the point how the router can tell to the corrdinator that it is a corrdinator. OR I missed something, or something is really wrong of the firmware that you made for slae. I am there to help you fix this, I have 3 weeks vacation from now so I can fully work on this but I need it fixed. It's not possible to re-pair, every device (30) everytime as I plan to have device that will be hard to reset when they will be installed ... I dont care of the money, If you tell me that there is another device that don't have the pan_id problem, I will bought two. But for me the best was slae as written on your website.
EDIT: I got news as some of my device just rejoin but some of them are showing this in log:
Zigbee2MQTT:warn 2021-07-12 12:20:13: Received message from unsupported device with Zigbee model 'undefined' and manufacturer name 'undefined'
Zigbee2MQTT:warn 2021-07-12 12:20:13: Please see: https://www.zigbee2mqtt.io/how_tos/how_to_support_new_devices.html.
Zigbee2MQTT:warn 2021-07-12 12:23:03: Received message from unsupported device with Zigbee model 'undefined' and manufacturer name 'undefined'
Zigbee2MQTT:warn 2021-07-12 12:23:03: Please see: https://www.zigbee2mqtt.io/how_tos/how_to_support_new_devices.html.
Zigbee2MQTT:info 2021-07-12 12:37:42: MQTT publish: topic 'zigbee2mqtt/remote_grimpe', payload '{"battery":100,"brightness":1,"counter":1,"linkquality":57,"update":{"state":"idle"},"update_available":false}'
Zigbee2MQTT:warn 2021-07-12 12:41:41: Received message from unsupported device with Zigbee model 'undefined' and manufacturer name 'undefined'
Zigbee2MQTT:warn 2021-07-12 12:41:41: Please see: https://www.zigbee2mqtt.io/how_tos/how_to_support_new_devices.html.
Zigbee2MQTT:warn 2021-07-12 12:41:56: Received message from unsupported device with Zigbee model 'undefined' and manufacturer name 'undefined'
But again, I changed nothing and my device was working before and succefully detected
OH and last but not least, I setup my other CC2531 as a sniffer to see what really going on. If you need traces, just ask.
I see philips dimmer coming back when I push the button, but still wondering why the philips bulb, tradfri bulb don't come. It's because of the database bug ?
And for a tradfri dimmer, I see this in the log:
Info MQTT publish: topic 'zigbee2mqtt/bridge/log', payload '{"message":"interview_started","meta":{"friendly_name":"remote_chambre"},"type":"pairing"}'
Error Failed to interview 'remote_chambre', device has not successfully been paired
@Koenkk are you able a day to debug what happend on the stick ? I'm quite sure it is something on the stick rather than on z2m
Did you try the procedure I described in https://github.com/Koenkk/zigbee2mqtt/issues/7761#issuecomment-877230153 ?
Did you try the procedure I described in #7761 (comment) ?
yes, I answered you on telegram
I have changed pan_id, channel, network_key. Added a hue bulb. Stop/restart z2m: Error: AREQ - ZDO - stateChangeInd after 60000ms Then it try to restart it again annnnd: guess what: Error: Coordinator failed to start, probably the panID is already in use, try a different panID or channel
I still think the flashed firmware is corrupt (like in https://github.com/Koenkk/zigbee2mqtt/issues/6302). This has nothing to do with the stick or z2m, but rather with the setup that the firmware has been flashed with. Can you reflash the firmware on a completely different setup or with Uniflash? (https://slae.sh/projects/cc2652/#flashing)
Yes I told you already that I reflash using Windows with python bootloader tool. Now I have upgraded to 1.20-1 latest version. I have restart and get the: Error: network commissioning timed out - most likely network with the same panId or extendedPanId already exists nearby. Again im asking is there a way to know that the device has been flashed successfully and correctly ? Because it seems to do a CRC32 after flash so I don't see why It can be wrong to be honest ... But ok I will try uniflash now. Did you read the telegram message ? EDIT: same error with Uniflash under windows: Error: Failed to connect to the adapter (Error: SRSP - SYS - ping after 6000ms) And now I get this everytime it boots
Can you add sha1/md5 of the firmware .hex file to the github ? Then we can be sure that our download is correct please.
EDIT2: The bootloader python script is not working, he Uniflash is not working, only the SmartRF Flash Programmer V2 is working. I am now able to switch between my 2 slae without any problem, if I re-flash it, it restart sucesssfully. I think it is mandatory to avoid such problem to remove from both SLAE and the flashing guide under zigbee2mqtt the python script as it is not working. I guess lot of people are not using it, otherwise I don't see why the haven't such problem
Good to hear you fixed the problem. In some rare cases (like this and #6302) the cc2538 bsl flasher fails. This can be checked by reading both firmwares back (cc2538-bsl.py -r
) you will see that the firmwares differ.
Sure. but it could be nice to understand why, or simply remove the documentation for this flasher. Also can you add to github the sha1 or md5 of each firmware ? Just to be sure that the actual downloaded file is the correct one. Thanks for help on the debugging. Unfortunately the -r option is timeout after a while. For me this flasher is definitely bugged.
Hello, still have problem with slaesh after updating HA/supervisor/z2m. Was working fine before. I tried reflashing, no luck. Problem come from 18.1.1 z2m version. Tested another slaesh, same problem. Latest version of home assistant on the supported Odroid. EDIT: Changing PAN_ID made it work, but then I lost all my device. Why I need to change PAN_ID if I didn't change anything in hardware/config and only upgraded HA/z2M ???? Here is the full log:
And I found a coordinator_backup but is it possible to know which pan_id there is inside ?
Back in the days I had a dongle migration between CC2531 and the slaesh, had to redo the whole network with new PAN_ID I guess. IS it possible that when upgrading the package, an old database/or config with old pan_id has been erased ? when doing snapshot, the database is taking as a backup ?
Inside the coordinator_backup.json I have this pan_id:
"value": [ 98, 26 ], "len": 2
And I guess as I tested another pan_id, the coordinator_backup.json should have this one right now ? Thus not compatible with the old database ?
And last question, why when doing a Snapshot with hassio, database and old important stuff are not backup up ???