Got a virtuozzo vps *NODE* that I leased and today, it locked up twice (stopped pinging out of the blue). First reboot went into fsck mode successfully completed anyway.
how can I trace the cause of the lockup? I'm thinking iptables, but I'm not experienced with iptables, any tips appreciated.
I want to make sure flushing iptables won't lock up the node again
/var/log/messages shows this:
Quote:
Oct 7 03:30:03 server kernel: Fatal resource shortage: numiptent, UB 108.
Oct 7 05:40:03 server last message repeated 2 times
Oct 7 05:40:03 server kernel: Fatal resource shortage: numiptent, UB 108.
Oct 7 12:02:01 server kernel: VPS: 107: stopped
Oct 7 12:06:23 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Down
Oct 7 12:06:26 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Up 10 Mbps Full Duplex
Oct 7 12:06:38 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Down
Oct 7 12:06:40 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Up 10 Mbps Full Duplex
Oct 7 12:06:44 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Down
Oct 7 12:38:51 server shutdown: shutting down for system halt
Oct 7 12:06:23 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Down
Oct 7 12:06:26 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Up 10 Mbps Full Duplex
Oct 7 12:06:38 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Down
Oct 7 12:06:40 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Up 10 Mbps Full Duplex
Oct 7 12:06:44 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Down
Oct 7 12:38:51 server shutdown: shutting down for system halt
|
It appears the main problem is:
Oct 7 18:46:56 server kernel: e1000: eth0: e1000_watchdog_task: NIC Link is Up 10 Mbps Full Duplex
Kernel: 2.6.9-023stab044.11-enterprise #1 SMP
O/S: CentOS release 4.3 (Final)
vps node locks up?
