Network server doesn't accept rejoin request

I have seen others have had the same issue, I read the answers but I can’t still fix the issue. I have nodes of a3ip , they are designed to request join periodically even if they are connected. it has been more than 6 months that they were working. but recently I saw there is no payload in device data and just join requests, I didn’t do anything and the nodes came back themselves. But it happened again and this time they didn’t come back. just join request and no join accept.
In device data, I see a few (one or two) validate dev-nonce errors so I used Clear Devnonce button but it didn’t work.
Beside , I checked that I have already dump file in /var/lib/redis/dump.rdb and as it is Debian / Ubuntu, persistence is enabled by default.
I have chirpstack v3.
gateway bridge on the gateway (kirlink)
I restarted the network and application server, but not the gateway and the nodes (as I don’t have access).
I need to bring the nodes back, I appreciate you answers , thank you.

It is always helpful if you can give the screenshots of the LoRaWAN frames and events.


below the photos for last messages of the sensor in device_error

below the photos for last messages of the sensor in device_join

the dev_addr culumn was different for each row (I don’t know if it is important)
I see more validate dev-nonce errors but I used Clear Devnonce button and it didn’t work.
I hope these photos are enough.

You can try to temporarily disable Frame Counter check in the device.

I tried but it didn’t work :frowning:

sorry, juste to ask if you have any suggestion that I can try. Thank you