Summary: | net-misc/dhcpcd-5.2.{8,9} gives no kernel routing table | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Peter Karlsson <peterk2> |
Component: | Current packages | Assignee: | William Hubbs <williamh> |
Status: | RESOLVED INVALID | ||
Severity: | normal | CC: | base-system, jordan.w.bradley, roy |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | x86 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | emerge --info |
Description
Peter Karlsson
2010-12-23 12:58:01 UTC
Created attachment 257869 [details]
emerge --info
Comment on attachment 257869 [details]
emerge --info
This is emerge --info with my working kernel/dhcpcd info (gentoo-sources <2.6.35 dhcpcd-4.0.15)
What output do you see from the init system? Anything in dmesg or the syslog? please post the output of the following dhcpcd -x dhcpcd -dB Then in another terminal, the routing table route -n (In reply to comment #3) > What output do you see from the init system? Anything in dmesg or the syslog? > I don't see anything in dmesg or syslog (/var/log/messages) related to dhcpcd unfortunately... The output from init does not look "suspicious": dhcpcd version 5.2.8 starting... rebinding lease 192.168.1.101 bla bla... However, the error may be on my side, see comment below. (In reply to comment #4) > please post the output of the following > dhcpcd -x > dhcpcd -dB > > Then in another terminal, the routing table > route -n Hm, this is embarassing but it seems the error is on my side. I have two network interfaces (eth0 and eth1). eth0 address is handled by dhcp and eth1 is static. So I do get a kernel routing table... but the order of the network interfaces is wrong, hence nameserver lookups goes to the wrong interface. With 4.0.15 I get a working kernel routing table (I just issue 'route'): Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 192.168.1.0 * 255.255.255.0 U 0 0 0 eth0 192.168.0.0 * 255.255.255.0 U 0 0 0 eth1 loopback * 255.0.0.0 U 0 0 0 lo default 192.168.1.1 0.0.0.0 UG 0 0 0 eth0 default server1 0.0.0.0 UG 1 0 0 eth1 (server1 = 192.168.0.1) With 5.2.{8,9} I assumed I didn't get a routing table because the output of 'route' (without the -n flag) displays just the headings (I assume it's trying to lookup the name of 'server1' which gets the wrong ip) so I assumed there's no kernel routing table. I should have been more patient. Anyway, this is what the routing table looks like with 5.2.8: Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 192.168.1.0 0.0.0.0 255.255.255.0 U 203 0 0 eth0 192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 eth1 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo 0.0.0.0 192.168.0.1 0.0.0.0 UG 0 0 0 eth1 0.0.0.0 192.168.1.1 0.0.0.0 UG 203 0 0 eth0 So it seems to be a configuration problem (as I thought from the beginning but haven't had the patience to resolve). Sorry for all the fuss. Should I close this as NOTABUG? Well, it's NOTABUG with dhcpcd :) So newer dhcpcd versions always use route metrics on Linux. It calculates the metric by the interface is wired, wireless, ppp and then interface number. This allows us to have >1 default route and the kernel will always use the lowest numbered route by preference. So to fix you need to set metric to zero in dhcpcd.conf OR set the metric in conf.d/net for your static interface to be above 203 generated by dhcpcd. Set it to something like 800. (In reply to comment #7) > Well, it's NOTABUG with dhcpcd :) Thank you so much for that info; I would probably have grown quite a few grey hairs otherwise... :-) Closing this bug as INVALID. Again, sorry for bogus bug reporting and I wish you and all gentoo devs a merry xmas and a happy new year! I added "metric 0" to /etc/dhcpcd.conf and installed 5.2.9, I still can't connect with networkmanager and dhcpcd takes a long time to start up. Does dbus autoload dhcpcd? I didn't add dhcpcd to the default runlevel. |