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
FS#2282 - The package odhcpd seems to be something wrong #7170
Comments
PussAzuki: It looks like I didn't save the dump file correctly. |
jow-: Please provide your network and firewall configuration |
PussAzuki: Things in the /etc/config/dhcp config dnsmasq config dhcp 'lan' config dhcp 'wan' config odhcpd 'odhcpd' I use a local dns resolver named Pcap-dnsproxy. Things changed in the /etc/config/network config interface 'lan' Things in the /etc/config/firewall config defaults config zone config zone config forwarding config rule config rule config rule config rule config rule config rule config rule config rule config rule option name 'Allow-ICMPv6-Input' config rule config rule config rule config ruleconfig rule config rule config include |
PussAzuki: The package dnsmasq-full seems to have the same problem, would it be a system error instead of a software error? |
PussAzuki: Is there anything I can offer? At present, it seems that dnsmasq-full and odhcpd/odhcpd-ipv6only have problems on my machine. With the former, the Android phone has only one internal network starting with fe80, and cannot connect to other ipv6 addresses. The latter got two public network addresses (one for SLAAC and one for non-SLAAC) and one for intranet addresses starting with fe80 on ipv6, and then it seems that the lifetime of 1800 seconds can't connect to other ipv6 addresses. I have been using another router to dial, and then this router is wirelessly bridged to reduce my relative distance to another router. |
PussAzuki: After many clean flashes, odhcpd metaphysically can renew the client's contract. But after restarting the odhcpd service, it can no longer renew the client. This is really a metaphysical bug. |
dedeckeh: Please attach logread traces of your findings. |
PussAzuki: Do I just change odhcpd's log level to 7 then reconnect the connection and use windump on Windows? If you can give me what exact command I need to use with windump, it should be better. |
PussAzuki: The windump started after I plugged the wire, and tcpdump started from 28 minutes. |
PussAzuki: This time it looks like there is no problem... |
dedeckeh: I cannot see anything wrong in the tcpdump; prefix 2001:470:d:1ca4::/64 is sent in RA prefix info option with the onlink and autonomous flags set as expected. The router lifetime is set to 1800 as expected with the other and managed flags set. Checking the dhcp config why are you setting ra_default to 1 ? |
PussAzuki: The file I uploaded is the one with no errors. It seems that everything is normal now. But if I use a wired link and then use a wireless connection on the same computer, it may not work properly again. But I can’t guarantee that it will be 100% recurring.A long time ago, whether I setted ra_default to 1, it wouldn't work properly, and I think it's a bug...now I need more time to find out when and how. |
PussAzuki:
Hello, this occurs on PSG1218 ver. A (MTK MIPS) on a fresh build from trunk.
I have used the 6in4 tunnel for a long time, and keep the same config for so long, so I want to have a clean flash, and it looks to go bad.
I lost the public network IPv6 default gateway and got an IPv6 default gateway for the intranet starting with fe80:
It seems that keeping the configuration unchanged for a long time may lead to some hidden problems. And they may only show up when you take a clean flash.
I have collected some information by logread and windump, and I hope it can do something.
The text was updated successfully, but these errors were encountered: