FreeBSD 9.0-p4 network unreachable after em0 state change UP -> DOWN -> UP

Hi,

seem to be having an issue where I can no longer remotely access my server after the network changes from UP -> DOWN -> UP, e.g

Code:
Nov 24 15:59:55 host kernel: em0: link state changed to DOWN
Nov 24 15:59:57 host kernel: em0: link state changed to UP

The machine has multiple IPs. Snippet of my rc.conf

Code:
# IPv4
ifconfig_em0="inet 142.4.xxx.xxx netmask 255.255.255.0 broadcast 142.4.xxx.255"
ifconfig_em0_alias0="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
ifconfig_em0_alias1="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
ifconfig_em0_alias2="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
ifconfig_em0_alias3="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
ifconfig_em0_alias4="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
ifconfig_em0_alias5="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
ifconfig_em0_alias6="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
ifconfig_em0_alias7="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
ifconfig_em0_alias8="inet x.x.x.x netmask 255.255.255.0 broadcast x.x.x.255"
defaultrouter="142.4.x.x"

# IPv6
ipv6_enable="YES"
ipv6_network_interface="em0"
ipv6_ifconfig_em0="2607:5300:xx prefixlen 64"
ipv6_ifconfig_em0_alias0="2607:5300:xx prefixlen 64"
ipv6_ifconfig_em0_alias1="2607:5300:xx prefixlen 64"
ipv6_ifconfig_em0_alias2="2607:5300:xx prefixlen 64"
ipv6_ifconfig_em0_alias3="2607:5300:xx prefixlen 64"
ipv6_ifconfig_em0_alias4="2607:5300:xx prefixlen 64"
ipv6_ifconfig_em0_alias5="2607:5300:xx prefixlen 64"
ipv6_ifconfig_em0_alias6="2607:5300:xx prefixlen 64"
ipv6_static_routes="ovhgw"
# ipv6_route_ovhgw="2607:5300:xx:xxff:ff:ff:ff:ff -prefixlen 128 -interface em0"
ipv6_route_ovhgw="-inet6 default 2607:5300:xx:xxff:ff:ff:ff:ff"

(yes I know ipv6 style has changed -- but that would not affect would it? I have not yet converted it)

This is a machine with OVH, so manually restarting routing every time this happens is not viable and requires a hard reboot.

Ideas?
 
It would be much more helpful if you had (physical or through some other means) access to your machine when this happens and show your arp -an , ifconfig -a and netstat -rn outputs. Moreover, if your aliased interfaces are on the same subnet as their parent interface (em0), it would be better if their subnet mask was 255.255.255.255 (which is not the source of your problem). Finally, if you have access on your machine when this happens, try arp -dan and see if the network comes to life after that.

Ah, and a final question: When does this UP->DOWN->UP take place? Because if it's due to you giving:

# /etc/rc.d/network restart

then the problem is that -most probably- your default gateway is emptied. You should start/restart your routing as well. So when restarting your network you should give something like this:

# /etc/rc.d/network restart && /etc/rc.d/routing restart

I hope this helps.
 
Back
Top