I have been using 5.10.41 kernel from development snapshots for a couple of months. The only issue is that IPv6 sometimes stops working. It happens like every 1 or 2 months, and I don’t know what triggers it.
I use 6in4 tunnel. Is there any way to troubleshoot it? What should I look for in the logs? What process might have stopped? Is there any known bug related to this?
Sure I could upgrade, but this image is otherwise stable, so I am a bit hesitating to install a more recent image.
i use 5.10.90 with my debian buster image on main-router (uptime 25 days), have no problem with ipv6 tunnel (using a HE ipv6 in ipv4 tunnel)
i guess the problem is that the update of your ipv4 to your 6in4 provides was not working…do you use a ip-up-script (i use one for ppp) for this update?
does it stops working on clients only (and still working on the host)? then maybe the radvd/dhcpv6 times out
I watched the IPv6 issue in the past couple of months. It happens only after an internet outage. Since such outage is very infrequent, the 6in4 tunnel stays up for months without breaking down.
Today it happened again at night. IPv4 and the DHCP clients recovered without any issue. But the 6in4 tunnel remained down. I rebooted the Banana Pi, but it did not help. So next action was a hard reset by unplugging the power supply. Then fixed the IPv6 issue.
Yes I use a cron job like this:
*/15 * * * * root /usr/bin/wget -q https://[username]:[password]@ipv4.tunnelbroker.net/nic/update?hostname=[tunnel ID] -o /dev/null
I suppose it will update the tunnel with my actual IPv4 address. At least it did work fine - until the provider outage.
I am monitoring the reachability of the 6in4-wan6 interface with Uptimerobot, not the clients. Actually it reported the IPv6 outage.