Network scripts / named not starting

[chielsen]

Yesterday my server was offline for serveral hours because the networkscripts did not start after a reboot (power failure).
I asked my dc to take a look, normally they dont give software support but the started the scripts so the server is online again.
As they don't give any further support, my question is how to keep this from happing again.

When i looked at the message, there were serveral messages that the service named and directadmin was down. I did some reboots but it didnt come up by its own. So what could have been wrong?

 

[GXX]

What do the logs show?

 

[chielsen]

In da log:

2008:03:10-12:39:11: ioctl can't find the server's ip address for eth0 : Cannot assign requested address
2008:03:10-12:39:11: The ip of this machine () does not match the ip in the license file.<br>

Hmm this seems to be it:

Mar 10 12:34:34 apophis named[1753]: starting BIND 9.2.4 -u named -t /var/named/chroot
Mar 10 12:34:34 apophis named[1753]: using 2 CPUs
Mar 10 12:34:34 apophis named[1753]: loading configuration from '/etc/named.conf'
Mar 10 12:34:34 apophis named[1753]: ifiter_ioctl.c:236: REQUIRE(iter->pos < (unsigned int) iter->ifc.ifc_len) failed
Mar 10 12:34:34 apophis named[1753]: exiting (due to assertion failure)
Mar 10 12:34:34 apophis named: named startup failed

ANd this is what happend to get it working again:

Mar 10 12:39:56 apophis network: Shutting down loopback interface: succeeded
Mar 10 12:39:56 apophis sysctl: net.ipv4.ip_forward = 0
Mar 10 12:39:56 apophis sysctl: net.ipv4.conf.default.rp_filter = 1
Mar 10 12:39:56 apophis sysctl: net.ipv4.conf.default.accept_source_route = 0
Mar 10 12:39:56 apophis sysctl: kernel.sysrq = 0
Mar 10 12:39:56 apophis sysctl: kernel.core_uses_pid = 1
Mar 10 12:39:56 apophis network: Setting network parameters: succeeded
Mar 10 12:39:56 apophis network: Bringing up loopback interface: succeeded
Mar 10 12:39:57 apophis kernel: ADDRCONF(NETDEV_UP): eth0: link is not ready
Mar 10 12:39:57 apophis kernel: e100: eth0: e100_watchdog: link up, 100Mbps, full-duplex
Mar 10 12:39:57 apophis kernel: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
Mar 10 12:39:58 apophis network: Bringing up interface eth0: succeeded
Mar 10 12:40:01 apophis directadmin: directadmin startup succeeded

 

[chielsen]

Anybody? Please?

 

[ju5t]

I would try and do some diagnostics on the NIC. Maybe the power failure broke something. Take the NIC out and replace it with a new one and try diagnostics on another server by logging in on-site or through KVM and perform the following command:

ethtool -t offline

Do not use this command on a production server; it will stop normal operations.