-
Notifications
You must be signed in to change notification settings - Fork 652
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Firmware 20221226 causes some NWK_TABLE_FULL errors with only 32 devices in network (19 routers) #418
Comments
I also got this error with 20221226 , I can’t remember why and when. 20221102 has been the most stable so far for me with sonoff 3.0 cc2652p usb stick |
Yup me too. Hardware: Edit: Edit 2: Edit 3: |
Same here, using ITead Sonoff 3.0 stick, forgot to mention but can tell I'm using launchpad fw. |
Same here, Sonoff Dongle Plus |
Not sure if @Koenkk is aware of this. |
|
Tried both cc-flasher (there's a Docker container with it, used it lot of times in the past) and the TI Flasher tool using Windows. Neither of them had an apparent problem. Using launchpad coordinator 20221226 firmware with CC2562P from ITead Sonoff Zigbee 3.0 USB Dongle. |
its far from stable... im runniung 141 devices around 120 routers devices and for weeks i have been getting this NWK Table Full error. |
@joaquinvacas that is strange indeed, I cannot come up with an immediate solution. When TI releases an new SDK I will compile and firmware and you can try it out. For now you can safely stay on |
Don't worry man, you already do a lot of work! Just reporting because there were no hw changes so only fw can make those errors. I'll roll back to 20220219 until then. |
I've just encountered the same problem. Out of a sudden, NWK_TABLE_FULL started randomly occurring, and the network became sluggish and unreliable. I correlate it with manually restarting a few devices on the network, which is something I've been doing many times before, without any issues. What helped, was re-pairing one of the routers and one of end devices (!). The network works as a charm again for over a day now. Coordinator CC2652P with CC1352P2_CC2652P_other_coordinator_20220219.zip + 9 routers + 4 end devices. |
for me Z-Stack_3.x.0_coordinator_20221226 firmware was beyond bad, my network dropped offline entirely. I go back to latest stable and repaired everything, cold not even salvage the network. |
Do you have any clue how I can find that version? I flashed to 20221226 and it's been a disaster. I seriously need to find some version of the firmware which works or my WAF is going to plummet. |
Looks like i made some typos here. I am on 20221102* |
@Koenkk How do you recommend debugging these firmware images after building + flashing them? Is there a way to get console output, diagnostic information, and/or "breadcrumbs" out of the TI microcontroller to figure out what events led up to a random failure (such as a NWK_TABLE_FULL error that happens after several days)? I just upgraded from 20220219 to 20221226 because on my mid-sized network (~40 devices) the ZC was constantly dropping reports. 20221226 seems to have fixed this for me, possibly due to increasing MAX_NEIGHBOR_ENTRIES. But if there are other issues on 20221226 I'd like to be able to debug them. I do have a J-Link probe and can order the TI dev boards if needed. |
@nickrbogdanov there are some generic docs about debugging: https://software-dl.ti.com/ccs/esd/documents/users_guide/ccs_debug-main.html, but I believe this error comes from the proprietary part of the firmware and can therefore not be debugged by us. |
i switch here as mine had been flaged as duplicate. Not better for my side with a downgrade of firmware. Still loosing devices. Too many devices ? I try to switch off the wifi 2.4Ghz completly. I change the USB cable to the SONOFF. I try without USB cable. Each times : |
Just a little feedback. |
Can you try the |
As we say in french : "Les grands esprits se rencontrent". I just push the 20221102 dev branch one hour ago. I'm repairing all end devices and i will see if it is better. |
Ok it's getting better.
|
@Koenkk same issue, trying to update firmware using TI flash Programmer. I get this message below?? any ideas??
|
Mayby it won't help but i use this for my sonoff: |
TI flash programmer doesn't work with the latest firmware (error: Unknown record type: 3.), you need to use, for example, the following python script: https://github.com/JelmerT/cc2538-bsl. For the Sonoff Dongle-p key, I used the following command line: Regarding the disconnections mentioned, I have another Sonoff Dongle-p key configured as a router, and it would become offline after a few hours. Reconnecting the key would solve the problem. |
Not so stable i've got 3 routers offline and not responding after a few hour. (and more NWK table fll errors). A lot of end devices are also inop even online (motion sensors). Edit : i'm trying to get some doc on this error but zigbee implementation is hard to understand for me. but it seems to have something about the number of routes in the mesh. |
Just to follow up: I've been running 20221226 for almost a month and it's been rock solid compared to 20220219. No problems at all with missing reports or dropped commands anymore. So it's possible that whatever knobs were changed in 20221226 vastly improved reliability on my network topology, while breaking it for others. It would be useful if the FW (or maybe a sniffer) could gather metrics on whatever anomalies are seen during operation, so that we can zero in on which settings might be to blame for the instability. Another option might be to make several builds using different parameters, to narrow down which change(s) caused the regression. |
Its not working with any firmware on either Sonoff Dongle i have. |
I really think there's something about the number of devices (no proofs just something i feel) cause sometimes a devices come back online but another one get offline. |
The main problem I had with the old 20220219 firmware was that the ZC would drop reports from a device for several minutes at a time, and then magically recover. This cycle would keep repeating every 10 minutes or so. This suggested to me that it had trouble juggling the number of (peer?) devices that I had in my network, so it could only maintain state for a subset of them at once. I have about 40 devices in my network. |
32 routers - 41 end devices |
I use this firmware 20221226 since it appears on master branch. I have practically any problem with this firmware but:
So far, anyway, these errors doesn't have bring specific problems. My network:
I wonder:
|
Hello, What should i do ? Changing my coordinator ? If yes which one ? Regards |
Sorry been meaning to jump on this thread. |
@guillaume042 growing the table more isn't the solution, it's already very big. Probably the ageing has to be changed, I will look into this a soon as TI releases their new SDK (which should be soon). |
Thank you ! :-) |
@Koenkk Is it possible to know how much the routing table is used ? |
@zen2 AFAIK not |
Let's continue in #439 |
This Philips Hue light always worked fine till this afternoon after I updated firmware to 20221226.
The text was updated successfully, but these errors were encountered: