[Gllug] Intermittent Network Loss on Centos Web Server

[JLMS]

On Tue, Sep 22, 2009 at 6:27 PM, George Siddiqui
<georgesiddiqui at gmail.com> wrote:
> I'm experiencing regular intermittent (10-20 per day)  network dropout
> (over SSH and HTTP) on a web server running Centos 5.3 and was
> wondering if someone here could suggest what I could try next.
>
> Much of the time  the network connection resumes itself - sometimes
> not.  When it doesn't (strangely enough) if I log the machine back on
> to itself as root via ssh - all network services resume for a number
> of hours till then it fails again.
>
> Things I have tried already:
>
> -Checked for duplicate IP's on network
> -Swapped Network cable
> -Swapping from eth0 to eth1
> -Updated the broadcom driver from v3.93 v3.99
> -Turned off the acpid service
> -Turned off the NetworkManager service
>
> It's using the tg3 driver for a broadcom nic.
>
> system logs shows the following but the eth1 is not ready messages
> does not correspond directly to any of the many network outages
>
> Sep 22 09:25:34 myserver kernel: eth1: Tigon3 [partno(BCM95703A30) rev
> 1002] (PCIX:100MHz:64-bit) MAC address*************
> Sep 22 09:25:34  myserver kernel: eth1: attached PHY is 5703
> (10/100/1000Base-T Ethernet) (WireSpeed[1])
> Sep 22 09:25:35 myserver kernel: eth1: RXcsums[1] LinkChgREG[0]
> MIirq[0] ASF[0] TSOcap[1]
> Sep 22 09:25:35 myserver kernel: eth1: dma_rwctrl[769c4000] dma_mask[64-bit]
>
> Sep 22 09:25:40 myserver kernel: process `sysctl' is using deprecated
> sysctl (syscall) net.ipv6.neigh.lo.base_reachable_time; Use
> net.ipv6.neigh.lo.base_reachable_time_ms instead.
> Sep 22 09:25:40 myserver kernel: ADDRCONF(NETDEV_UP): eth1: link is not ready
> Sep 22 09:25:40 myserver kernel: tg3: eth1: Link is up at 1000 Mbps,
> full duplex.
> Sep 22 09:25:40 myserver kernel: tg3: eth1: Flow control is on for TX
> and on for RX.
>
> [root at myserver ~]# ifconfig -a
>
> eth1      Link encap:Ethernet  HWaddr ***********
>          inet addr:*********** Bcast:***********  Mask:***********
>          inet6 addr: fe80::20d:60ff:fe1c:22d1/64 Scope:Link
>          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>          RX packets:176071 errors:0 dropped:0 overruns:0 frame:0
>          TX packets:24085 errors:0 dropped:0 overruns:0 carrier:0
>          collisions:0 txqueuelen:1000
>          RX bytes:17341622 (16.5 MiB)  TX bytes:18926143 (18.0 MiB)
>          Interrupt:177
>


Is the port in your switch also in full duplex mode?

Have you tried connecting the machine to a different switch/router?
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug