Losing internet access sounds like a configuration problem, an ISP doing things or a combination of both...
I only had trouble with my AX3000T because of overloading the ram with blocklists. Since reducing the blocklists it has been rock solid, internet-connection-wise.
Open a topic with your problem and all the details of your setup.
hey Alexq, I tried this MiWifi Repair Tool procedure to get back to stock but although the procedure finished, after reboot it goes back to original state. Did I do something wrong?
Try one more time via MiWifi Repair Tool (I assume you didn't install a custom OpenWrt U-boot bootloader and have the ootb Xiaomi bootloader on your rd23).
If not, you can try to proceed with the "*** U-Boot Boot Menu ***" and select point 8 (8. Load image) and follow the Debricking instructions from the Wiki.
Also, you can try installing OpenWrt again via the UART method (just start from step 5).
update:
Also, make sure that the downloaded OpenWrt files contain the same SHA256 checksum as specified on the download page (ust to ensure that the downloaded file is not corrupted).
A Chinese router with firmware 1.0.84 arrived. Today I have time to try the firmware. I have access to UART, I know how to use it, but it is at my friend's place for now, I can take it tomorrow. So if you don't want to wait until tomorrow, offer those that have less chance of bricking the router)))
yes, it does
ESMT NAND == supported by 23.05.5
Winbond NAND == supported by Snapshot only
AN8855 switch == not supported neither 23.05.5 nor Snapshot (at least for now).
There are lots of these topics with the same problem, and no proper solutions have been found. Some of them got the problem solved by updating to some new version, others misconfiguration, and others had no solution.
As you said, your version is working very well.. but then I ask: how many VLANs do you have? How many Firewall rules? Are you working with static Leases? Are you using more than one router and using the WLAN roaming?
This is my scenario here, and the worst thing, sometimes I have this loss 5 times a day, sometimes two in less than 30 minutes. Sometimes I didn't experience this issue for weeks.
That's why I just want to keep a more tested version like the stable one.
What do you think? Is it Worth it to open a new topic even before changing?
So far, there hasn't been any evidence from anyone that RD23 received a Winbond NAND chip or an AN8855 switch... But this could happen someday (so, please verify the NAND chip before flashing)
So, until the RD23 (Global version) comes with an EMST NAND chip, it's supported by both Stable 23.05.5 and Snapshot.
do we know why a linux kernel change needs to be implemented before the winbond nand chips are supported? what is so different about these nand chips? I would expect there to be a standard manufacturers follow when manufacturing nand chips so these issues don't happen when oems change part suppliers?
We know, different partition tables, which is normal, just these partitions need to be added to the firmware so that it knows at what addresses to flash itself correctly, so that the bootloader can load it correctly. This has actually already been done in SNAPSHOT. All that remains is to add a driver for the new network switch, otherwise the router turns into a regular switch))).
So what's the problem? I read not long ago that when assembling the firmware you need to add only one parameter and it was written which one. But unfortunately I'm not a programmer.
But I could check this with ready-made firmware))) Which is what I did. And none of the proposed firmware works properly. Which is what I wrote a little higher.