Replies: 56 comments 242 replies
-
|
I will try it right away :-) Btw. I got an spare slae.sh cc2652rb stick laying around I would setup as an sniffer and have it running even before I do the upgrade. Ive downloaded the TI sniffer firmware, but can't find any hex file matching that model. Is it even possible to use it as a sniffer or is there another firmware I should use?
UPDATE: I can see I need the Packet Sniffer 2, but that's locked behind a "paywall" and no matter what I do, I'm denied access, so I guess I'm unable to setup the sniffer :-( |
Beta Was this translation helpful? Give feedback.
-
|
You could use the conbee2 with the official sniffer firmware from Dresden Electronic and (zshark) Wireshark? In case this could help observing the misbehaving of the network. |
Beta Was this translation helpful? Give feedback.
-
|
unfortunately my conbee2 stick is dead, so its no good :-( |
Beta Was this translation helpful? Give feedback.
-
|
How could you kill a conbee2? The bootloader is static and can't be changed do you have flashed the firmware again on a Linux machin ? |
Beta Was this translation helpful? Give feedback.
-
|
My cat had too much fun with it :-D |
Beta Was this translation helpful? Give feedback.
-
|
That's sad and funny at the same time. 😳 |
Beta Was this translation helpful? Give feedback.
-
|
I have upgraded to 20231112 now and will get back in case of issues! My environment has been quite stable though even with the releases where some others have had issues. Small issues observed with latest few test releases, a device or two that has been disconnected but power cycle the IKEA bulb has put them online again without any need for repairing. The only thing which I believe has been unnormal (but infrequent) is unrequested restarts of z2m when trying to pair several lights at the same time, such as when pairing a set of new lights in a fixture so all announce them self at the same time. CC2652R (zzh) |
Beta Was this translation helpful? Give feedback.
-
This error started showing up in the log
|
Beta Was this translation helpful? Give feedback.
-
|
Yeah, somehow I got the same problems when returning to 20231127, but when I went all the way back to 20230507, the network came alive again :-) |
Beta Was this translation helpful? Give feedback.
-
|
@smulle48 does 20231201 work now? |
Beta Was this translation helpful? Give feedback.
-
|
@Koenkk No unfortunately not. I have attached the log from the startup of the 20231201 firmware |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
Happy to provide my feedback on 20231203, but I need a version without the tclk table limited to 100. |
Beta Was this translation helpful? Give feedback.
-
Should I re-pair them, or try |
Beta Was this translation helpful? Give feedback.
-
|
I've confirmed 
|
Beta Was this translation helpful? Give feedback.
-
|
Add me to the “20230922 is the best” list. I had a few devices I had to re-pair in the day or two after switching to it. However, I’ve now had 9 days of no adapter crashes or dropped devices. |
Beta Was this translation helpful? Give feedback.
-
|
Is there a way I can determine if I've hit the maximum devices my coordinator can support? I've added two devices for Christmas lights, going from 107 to 109, and in that time my network has become unstable again. There's nothing in the logs - just devices not responding or lagging. I'm thinking of ordering an additional coordinator and splitting part of my network, but if there is anything that would be helpful to test let me know. |
Beta Was this translation helpful? Give feedback.
-
|
its not as simple as a max number of devices for a certain coordinator/firmware, it highly depends on the amount of traffic on the network |
Beta Was this translation helpful? Give feedback.
-
|
Thanks, that was useful advice. With this in mind, I removed availability checks from a few bulbs I had on hard-wired switches (they are end devices, not routers, and for guests and kids there's value in knowing the light is off). I've now gone 3 weeks on |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
20231112 went fine excepted some ping warnings (~5 per days). I had to move my whole docker automation setup on a NUC as my sd card in my pi4 just stop working. For approximately a week, I still had my ping warnings but some light were slow especially when not used for a long time. I then restart my fresh pi (which is on a ssd now) and restarted my docker automation. It was a total nightmare...nothing was working, tons of errors. I decided to flash 20231111, same behavior. So I tried 20230507, no luck as well. I went on 20220219 and everything came back, no errors, no ping warnings. I then moved on 20221226 which went fine for days. A new try on 20230507 and everything was fine for days as well (some ping warnings came but only 2...) I am now on 20230922 which is working fine and I'm going to stay on this one (no ping warnings/error since it is up). After closely reading the discussion on TI forum, it is for me the best choice as something seems broken in SDK 6.20 and higher (20230922 in based on SDK 6.10). |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
Looks like errors are popping back up but appear different: |
Beta Was this translation helpful? Give feedback.
-
|
And the network itself? Some drop outs? |
Beta Was this translation helpful? Give feedback.
-
|
Hard to say as concurrently there was a bug introduced in Node Red (now resolved) that broke a lot of stuff so the malfunctions I observed were likely due to NR and not Zigbee. I did not change firmware so I will keep an eye on it. |
Beta Was this translation helpful? Give feedback.
-
|
Exactly what a was thinking. Thanks for your information. |
Beta Was this translation helpful? Give feedback.
-
|
Other issue was corrected but my logs are still packed with zigbee timeouts, etc. Any other firmware I should try? |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
Had 20231112 running for several weeks. I only had trouble with SONOFF SNZB-02D which had trouble with connection/paring and would not update anymore. |
Beta Was this translation helpful? Give feedback.
-
|
Running 20231111 for more then a week now. No problem/error what so ever. It even looks that the network is slightly faster. |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
Amount of devices in network: 87 20221226 worked fine: Pushed past this this version because of ongoing issues with Danfoss TRVs. I've been running 20230716 since release and found devices unresponsive on no more than a few occasions (and never a TRV). I'm back trying new versions because the TRVs and some light switches are now failing to respond. The recent changes to my network are: 20231111 feedback: Worked for several days, then significant number of errors failing update Danfoss TRV external temperatures, plus at least one MOES Dimmer switch unresponsive. I now have debugging turned on and will forward logs next time it falls apart. My assumption is that the same problems are surfacing here as with 20230716 and my next troubleshooting step will be to disable Adaptive Lighting. 20231112 feedback: Not tried as others seem to have difficulty - I need a fairly stable network. 20231115: Failed to launch Z2M after flashing with following error, so flashed back to 20231111
|
Beta Was this translation helpful? Give feedback.
-
|
Network had a meltdown overnight with majority of devices ceasing to work... so logs attached as promised. Restarting z2m only partially helped, but a power-down reboot of the pi brought the network back as it was before. It ran OK for about 8 days, was playing up a bit and then kaboom on the 10th day. I hope I managed to enable debug logging OK and attached the correct files. First file shows the 9/10 days of running, second file is after the soft restart (at which point I removed the USB stick manually then did a hard reboot). |
Beta Was this translation helpful? Give feedback.
-
|
20231112 no Here 20231112 not working. End devices stops responding and magicly sometime comes alive again. Router devices sometimes have a slow respons time. Going back to 20230716 this seems to be the most stable one for me. Could it be that not only the firmware is causing irrathic beheavier on the network but certain devices do to. |
Beta Was this translation helpful? Give feedback.
-
|
Number of devices on the network: 74 (house) Bridge software (Zigbee2MQTT/ZHA): 20221226 worked fine: don't know was on a conbee2 never had problems 20230507 worked well: no 20230922 feedback: Testing since 0922 it works like a charm no dropouts, no network crashes, all devices since firmware flash operational, thank god because my wife would kill me ;) same for my sister's house 82 devices, I think the only thing that's disabled is the updates on the firmware site? Is this possible? 20230923 feedback: no Test on a secondary network 23 devices. 20231112 no 20231111 not tested yet Cheers 🥂. |
Beta Was this translation helpful? Give feedback.
-
The error is this , about router: Failed to ping '0x00124b0024c2b4ca' (attempt 1/1, Read 0x00124b0024c2b4ca/8 genBasic(["zclVersion"], {"sendWhen":"immediate","timeout":10000,"disableResponse":false,"disableRecovery":true,"disableDefaultResponse":true,"direction":0,"srcEndpoint":null,"reservedBits":0,"manufacturerCode":null,"transactionSequenceNumber":null,"writeUndiv":false}) failed (Timeout - 60997 - 8 - 9 - 0 - 1 after 10000ms)) I continue to keep it for a few more days ;) .... upodate Most error to read LQI when show map (only router device) ;) Error 2023-11-24 12:56:15Publish 'get' 'transmit_power' to '0x00124b0024c93d58' failed: 'Error: Read 0x00124b0024c93d58/8 genBasic([4919], {"sendWhen":"immediate","timeout":10000,"disableResponse":false,"disableRecovery":false,"disableDefaultResponse":true,"direction":0,"srcEndpoint":null,"reservedBits":0,"manufacturerCode":null,"transactionSequenceNumber":null,"writeUndiv":false}) failed (Timeout - 12325 - 8 - 167 - 0 - 1 after 10000ms)' more error when change state ... Error 2023-11-24 13:16:02Publish 'set' 'state' to '0xa4c138300227844d' failed: 'Error: Command 0xa4c138300227844d/1 genOnOff.off({}, {"sendWhen":"immediate","timeout":10000,"disableResponse":false,"disableRecovery":false,"disableDefaultResponse":false,"direction":0,"srcEndpoint":null,"reservedBits":0,"manufacturerCode":null,"transactionSequenceNumber":null,"writeUndiv":false}) failed (Timeout - 35554 - 1 - 13 - 6 - 11 after 10000ms)' |
Beta Was this translation helpful? Give feedback.
-
|
Hi, sorry for long post, but I have the same issues as you, always timeout for some device. None of the firmwares worked reliable for me, even 20221226 that seamed to work fine for some time on Sonoff Dongle-P and Z2M, started to drop one or two devices daily, every day. I was repairing them, and next day the same issue. I don't know what caused this, I didn't added any new device nor changed anything. It just started to happen. My network is not so large, it has ca. 60 routers and 27 end devices. And with all due respect to Koenekk and his work, this is a complet mess for me, it has nothing to do with smart home or reliable automations. A day ago I've installed 20231112, to check if it will resolve my issues, of course it did not. Slowdowns were noticable already after 30 seconds, to a point where all devices were totally unresponsive. I was going back even to 2021 firmwares, always the same issues, NWK Table full, timeout, SRSP - ZDO, etc... Yesterday when going back to 20221226 I needed to flash Sonoff dongle three times using Powershell before it started to work somehow, first two times caused slowdowns, |
Beta Was this translation helpful? Give feedback.
-
|
today i came home and the light didnt work. |
Beta Was this translation helpful? Give feedback.
-
|
According to the Zigbee2mqtt documentations, the home assistant dongle is still considered as experimental (https://www.zigbee2mqtt.io/guide/adapters/#experimental). Just some tips you may have tried already.
|
Beta Was this translation helpful? Give feedback.
-
|
As far as I have tested, the Blue Sky home assistant is largely stable, the only pain is getting new devices up and running, which was such a pain for me, but once you have all your devices online it was mostly OK. I think the conbee2 and CC2652p are actually the best choice for zigbee2mqtt until we find out what's going on. I've actually been using the 0922 since 0922 without any problems. |
Beta Was this translation helpful? Give feedback.
-
|
I came from the Conbee2 Stick that's a nice piece of Hardware. But fast reaction times i only get with there deconz/phoscon software. And you have to separate the Stick from your PC/Raspi etc. with a long USB-Cable to minimize interferences. Now i use the smlight SLZB-06 (without M) because it's POE Powered and i can place it anywhere who is Ethernet. And it's rock solid and fast reaction times with the actual Firmware 2.0.8-dev and the Zigbee Firmware 20230922 from here. (edit some typo) |
Beta Was this translation helpful? Give feedback.
-
|
after 3 weeks running fine with 20230922, started again NWK_TABLE_FULL :-( Solved with a full power off power on of the entire VM. Including a removal of the sonoff-p stick and reseat @Koenkk are you aware of any other coordinator running stable with z2m? . I'm available to buy something new, and eventually reconnect all the 60 devices. error 2023-11-27 18:32:28Failed to configure 'Condizionatore Marco energia', attempt 1 (Error: Read 0xa4c138b9b11ecff7/1 genBasic(["manufacturerName","zclVersion","appVersion","modelId","powerSource",65534], {"sendWhen":"immediate","timeout":10000,"disableResponse":false,"disableRecovery":false,"disableDefaultResponse":true,"direction":0,"srcEndpoint":null,"reservedBits":0,"manufacturerCode":null,"transactionSequenceNumber":null,"writeUndiv":false}) failed (SREQ '--> ZDO - extRouteDisc - {"dstAddr":3066,"options":0,"radius":30}' failed with status '(0xc7: NWK_TABLE_FULL)' (expected '(0x00: SUCCESS)')) at Object.func (/app/node_modules/zigbee-herdsman/src/adapter/z-stack/znp/znp.ts:322:27) at Queue.executeNext (/app/node_modules/zigbee-herdsman/src/utils/queue.ts:32:32)) |
Beta Was this translation helpful? Give feedback.
-
|
Alright, I had slowdowns in 1112. I unfortunately have no logs there...we had a power outage (a few actually), and that reset the network and corrupted my log (battery backup apparently dead...) I then tried the 1127 I saw posted and I had the slowdowns/devices dropping off line. Logs from Home Assistant: |
Beta Was this translation helpful? Give feedback.
-
|
Are we sure when all devices go offline it's a firmware problem only? |
Beta Was this translation helpful? Give feedback.
-
|
Not sure why you have these problems. Version 20230922, mentioned above by several people, works solid and with no errors for most people. Have you read the many posts on zigbee network reliability, regarding USB extenders, wifi interference, usb interference? |
Beta Was this translation helpful? Give feedback.
-
|
I discovered this thread only few days ago so I couldn't partecipate in testing those firmwares.
Regarding interferences, I have an USB extender (I had this one since Deconz), wifi channels at home are 1 and 6 while zigbee network is 25. Probably everything I wrote here is useless for this thread but it could be useful to understand I'm not a passive user. :-) P.S. I think people who have a stable network should write which devices have. I suppose there are devices that disturb the network. |
Beta Was this translation helpful? Give feedback.
-
|
My suggestion would be to open a new thread on the home assistant forum under zigbee. This post is for the firmware testing. |
Beta Was this translation helpful? Give feedback.
-
|
Hey Hey 👋 |
Beta Was this translation helpful? Give feedback.
-
|
I also got the instablility issue with 20230507 and backed to 20221226 for a long time. Recently, when our devices are grown up more than 80 router (including 15 zigbee presense sensors), the zigbee network is not stable as before even with 20221226. There were some failed LQI error frequently as below:
I will continue to try the version 20230922 and report the result. |
Beta Was this translation helpful? Give feedback.
-
If possible, provide herdsman debug log or even sniffing logs from the moment everything works fine until it fails. |
Beta Was this translation helpful? Give feedback.
-
|
Maybe not the best solution, but since I've made a cronjob restarting zigbee2mqtt every third day, I have had zero issues. First I had it set to two days, but that still gave a lot of errors, so I guess the network takes more the 48 hours to settle, but less then 72 hours. I will try to increase the time between the restarts to see if it affects anything |
Beta Was this translation helpful? Give feedback.
-
|
Restart's really just a workaround! |
Beta Was this translation helpful? Give feedback.
-
|
Same for me, sometimes some devices simply disappear from Z2M completly with no reason. So restarting is only the last option |
Beta Was this translation helpful? Give feedback.
-
|
For anyone interested how I was able to stabilize my zigbee network after months of issues, also described in this topic, here is my absurdly simple solution. For me most stable version was 20221226, but even then I've had many issues, timeouts, not responding devices, slowdowns, etc., and nothing helped. Despite using RPi4B and Sonoff Dongle P on a extension cable, using 25th network channel, additional Dongle P zigbee router and so on, issues persisted and I was close to burn everything to the ground. In a desperate move (thinking that this will not change anything) I've decided to move one of my Asus AI Mesh nodes much, much further away from my Dongle P and a miracle happened. For over a month I have 100% stable network, without ANY not responding devices, timeouts or other daily issues. I also cannot believe it, but nothing else was changed, nothnig. Now I'm finally happy to use my zigbee devices. |
Beta Was this translation helpful? Give feedback.
-
|
Different room, ca. 10 meters, so also wall between them. |
Beta Was this translation helpful? Give feedback.
-
|
how close were they originally standing? |
Beta Was this translation helpful? Give feedback.
-
|
Ca. 1,5 - 2 meters |
Beta Was this translation helpful? Give feedback.
-
|
it's really weird. are the asus mesh nodes connected over the air? if so, have you tried scanning the air with e.g. wiifi explorer? such devices often communicate with each other on channels other than those on which they communicatу with clients. |
Beta Was this translation helpful? Give feedback.
-
|
No, the nodes are connected via lan backhaul. It is strange, maybe the signal from it was too strong, maybe it's due to many multidirectional antennas, I simply don't know. No router settings were changed, I simply moved it away. I know that you need to believe my word, but something caused the interference, and from day one when I moved the node, issues were gone, just like that. I've waited several weeks to check if this is just few days of luck, if the issues will come back, but everything is working as never before. For the ones that want to know which Asus mesh system it is, it's ZenWiFi Pro ET12. |
Beta Was this translation helpful? Give feedback.
-
|
I wish I had an asus mi I could do that to. Same symptoms no asus.
Sent via RFC 1149
Get some Carrier Pigeons<https://en.m.wikipedia.org/wiki/IP_over_Avian_Carriers>...
…________________________________
From: SmolikOne ***@***.***>
Sent: Sunday, February 4, 2024 6:00:53 PM
To: Koenkk/Z-Stack-firmware ***@***.***>
Cc: johnlento ***@***.***>; Comment ***@***.***>
Subject: Re: [Koenkk/Z-Stack-firmware] Z-Stack_3.x.0 coordinator 20231111/20231112 feedback (Discussion #483)
It was on automatic, so it varied from time to time.
—
Reply to this email directly, view it on GitHub<#483 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AGHSRT5Q634NAXDAEK4CGOLYSAHKLAVCNFSM6AAAAAA7HDUV26VHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DGNRTGY4DO>.
You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Maybe this information helps: I had some battery operated devices, with dead batteries (vibration sensors). Zigbee2MQTT was crashing every couple of days, random period. After fully removing the devices, the network is now up and running for 12 days without a problem. `Zigbee2MQTT version Coordinator type Coordinator revision Coordinator IEEE Address Frontend version Zigbee-herdsman-converters version Zigbee-herdsman version Stats |
Beta Was this translation helpful? Give feedback.
-
|
I am currently using version 20230507, but I am also encountering issues where devices go offline or repeaters go offline. Which version should I opt for? Based on the discussion, I am unable to determine a working version. |
Beta Was this translation helpful? Give feedback.
-
|
Basically use the 20221226 router fw that's made with the same sdk version |
Beta Was this translation helpful? Give feedback.
-
|
came from from 20221226 and upgraded to 20230922, after a few days I lost two of my Danfoss Ally Thermostats whereas it used to be rock solid previously. |
Beta Was this translation helpful? Give feedback.
-
|
I will provide a 6.10 fw (including p7) soon, currently doing some testing |
Beta Was this translation helpful? Give feedback.
-
Whats the difference between the current released router 20221102. and your suggested 20221226 ? |
Beta Was this translation helpful? Give feedback.
-
|
any update on a new router fw? |
Beta Was this translation helpful? Give feedback.
-
|
Came from 20230507 to 20230922. |
Beta Was this translation helpful? Give feedback.
-
|
The last time I worked with Ledvance in Munich they didn't even know their products were Zigbee or Wi-Fi, it was a mess. The company is getting too big too fast. |
Beta Was this translation helpful? Give feedback.
-
|
Nevertheless they are selling new zigbee products under VIVARES series |
Beta Was this translation helpful? Give feedback.
-
|
Are you sure the devices are offline? You mention you can "wake" then using the dev console? If they are offline, then you can not do anything besides "turn off/turn on" on the devices. Might it be they work perfectly, however show offline in Z2M, as they do not report every hour? And do not respond to the "live beat". I have some Osram, they work perfectly, however only report every 6 hours. You need to set this up in Z2M device file. Might be true for a lot of you other devices? Like this: |
Beta Was this translation helpful? Give feedback.
-
|
If they are offline in Z2M means they are "greyed" in HA so they aren't available.
|
Beta Was this translation helpful? Give feedback.
-
|
Availability is not a Zigbee network feature, however a Z2M feature, looking for when a device did the last reporting. I have some devices which will never report, hence I have to disable the availability reporting for those devices in Z2M. No matter if a device is reported as "offline" it will have the same behavior. This fits your point 2 and 3 above. For your point 1 and 4, this is different issues. |
Beta Was this translation helpful? Give feedback.
-
|
I have so many devices very close each others that zigbee mesh should "cure" eventual problems of routing. |
Beta Was this translation helpful? Give feedback.
-
|
I had a bunch off zbmini's originally in my zigbee2mqtt network and indeed they work irrathicly. |
Beta Was this translation helpful? Give feedback.
-
|
Yes I experienced the same thing, having 2 different zigbee networks is not so good, but for a while it was also my best take, I know from my own experience what's good and not good for me. Also that I have more cases where I admin some zigbee networks. |
Beta Was this translation helpful? Give feedback.
-
|
@Koenkk: I have been following these threads since the beginning. |
Beta Was this translation helpful? Give feedback.
-
|
This would be one of the best things and approach, maybe some z2m feature to send metric information? Just as an option (opt-in) so that we could collect some very interesting information? |
Beta Was this translation helpful? Give feedback.
-
|
You could use a sniffer and wireshark and process the logging. |
Beta Was this translation helpful? Give feedback.
-
|
I'm agree an network-debug option to opt-in in configuration will fit it. |
Beta Was this translation helpful? Give feedback.
-
|
The sniffer way is too much complex:
So it is not so easy that something integrated to Z2M that users could simply copy/paste results. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
A tool that discovers who is the guilty in the mesh would be very useful. |
Beta Was this translation helpful? Give feedback.
-
|
There will be too many parameters to give a 'quilty' to a device. |
Beta Was this translation helpful? Give feedback.
-
|
@Koenkk : I use UZG-01 zigbee coordinator with p7 (fw rev. 20230507). Even though I only have now around 30 Zigbee devices in use, I keep having the 'table full' problem. This occurs sporadically, sometimes more and sometimes less, and I haven't been able to identify the trigger yet. Can we predict when the new firmware for the p7 will be ready and will the problem be solved then? Thank you very much for the feedback. |
Beta Was this translation helpful? Give feedback.
-
|
I've created a new firmware, let's continue in #496 |
Beta Was this translation helpful? Give feedback.
and others
-
Update: a new test firmware has been released: #496
Since some users are having issues with the
20230507firmware which they didn't have with20221226, I've compiled 2 new test firmwares. This is a continuation of #474 and #478.Download: 20231111_20231112.zip
First start by testing
20231112, if that doesn't work try20231111. If20231112works fine, there is no need to try20231111When providing feedback copy and comment the template below:
Compared to 20230922/20230923 these contain the improvements based on this TI topic
Beta Was this translation helpful? Give feedback.
All reactions