 | | | strange network problem | strange network problem 2004-02-10 - By Xu Congyuan, Patrick
Back
crucificator wrote:
> Xu Congyuan, Patrick wrote:
>
> > crucificator wrote:
> >
> > > Xu Congyuan, Patrick wrote:
> > >
> > > > I am upgrading from redhat 8 to redhat el es 3. I have problem
> > > > with the netowrk.
> > > >
> > > > This server has two network cards installed. One is connected to
> > > > our internal network. The other one is connected to internet.
> > > >
> > > > The problem is with the external network.
> > > >
> > > > When the server is up for a while, the external network is down.
> > > > It means it cannot be accessed via its external ip address. The
> > > > internal network is working fine.
> > > >
> > > > I found out the external network can be resumed by "ifdown eth1;
> > > > ifup eth1 ", which means to restart the network card.
> > > >
> > > > When the external network is down, I run "ifconfig ", and the result
> > > > shows eth1 is "UP " and working. I really cannot figure out what
> > > > caused this problem. Now what I can do is to restart the "eth1 "
> > > > every 10 minutes.
> > > >
> > > > Any suggestion?
> > > >
> > > > The result of "ifconfig " is shown below.
> > > >
> > > > eth0 Link encap:Ethernet HWaddr 00:03:47:F1:94:5A
> > > > inet addr:192.168.42.248 Bcast:192.168.42.255
> > > > Mask:255.255.255.0
> > > > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> > > > RX packets:22271 errors:0 dropped:0 overruns:0 frame:0
> > > > TX packets:9294 errors:0 dropped:0 overruns:0 carrier:0
> > > > collisions:0 txqueuelen:1000
> > > > RX bytes:3229764 (3.0 Mb) TX bytes:2858019 (2.7 Mb)
> > > > Interrupt:10 Base address:0x1400 Memory:fe7e0000-fe7e0038
> > > >
> > > > eth1 Link encap:Ethernet HWaddr 00:03:47:F1:94:5C
> > > > inet addr:202.73.42.108 Bcast:202.73.42.111
> > > > Mask:255.255.255.248
> > > > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> > > > RX packets:20886 errors:0 dropped:0 overruns:0 frame:0
> > > > TX packets:12747 errors:0 dropped:0 overruns:0 carrier:0
> > > > collisions:4 txqueuelen:1000
> > > > RX bytes:3119008 (2.9 Mb) TX bytes:5289086 (5.0 Mb)
> > > > Interrupt:5 Base address:0x1440 Memory:fe780000-fe780038
> > > >
> > > > lo Link encap:Local Loopback
> > > > inet addr:127.0.0.1 Mask:255.0.0.0
> > > > UP LOOPBACK RUNNING MTU:16436 Metric:1
> > > > RX packets:200665 errors:0 dropped:0 overruns:0 frame:0
> > > > TX packets:200665 errors:0 dropped:0 overruns:0 carrier:0
> > > > collisions:0 txqueuelen:0
> > > > RX bytes:19471584 (18.5 Mb) TX bytes:19471584 (18.5 Mb)
> > > >
> > > >
> > > >
> > > >
> > > Just a thought... Maybe APM brings down the card?
> > > Look in the logs maybe you can find something strange. Although if
> > > you say that the OS reports the interface as up...
> > > What exactly means "a while "?
> > > Ah, just saw... 10 minutes. If it 's exactly 10 minutes and occurs
> > > constantly then you must realise that there is something
> > > configurable. Anyway give us the output from ifconfig eth1 when your
> > > connection is down and the exact occurence of the spookie thing :).
> > >
> > >
> > >
> > "A while " can be 1 minute, 10 minutes, etc. It is not fixed. As I
> > have said, the result of ifconfig eth1 is showing it is "UP " when the
> > external network is down.
>
>
> Yes but do you still have it configured (IP, mask, gw, etc)?
>
> >
> >
> > Maybe I should look into the system log file.
>
>
> Maybe you should... :)
>
> >
> >
> > Btw, this is my company 's web server. So I cannot afford to let it
> > down. Currently I did a cron job to restart the eth1 every 10 minutes.
> >
> >
> Brute force, ha? :)
>
> As far as I am concerned unless you get more specific there could be a
> lot of problems:
> I see you have 4 collisions on eth1. What device do you have ,
> topologically speaking, between your server and your ISP? Maybe you
> have a switch that died on you... I 'm not sure what kind of device you
> might have that is not providing L3 filtering.
> Try to change your eth1 card.
> Try lower your interface to 10Mbps halfduplex and see if you get the
> same result.
> Stop shorewall and see if this still happens.
> Check at boot time for other devices that might use the same IRQ as
> eth0 (IRQ5).
I checked the syslog. The eth1 is actually running at 10Mbps
halfduplex! How come...
Feb 9 14:52:33 www kernel: PCI: Found IRQ 10 for device 00:0f.2
Feb 9 14:52:33 www kernel: PCI: Sharing IRQ 10 with 00:0f.3
Feb 9 14:52:34 www kernel: PCI: Found IRQ 5 for device 00:04.0
Feb 9 14:52:34 www kernel: e100: eth1: Intel(R) PRO/100 Network Connection
Feb 9 14:52:34 www kernel: Hardware receive checksums enabled
Feb 9 14:52:34 www kernel: cpu cycle saver enabled
Feb 9 14:52:34 www kernel:
Feb 9 14:52:34 www kernel: ip_tables: (C) 2000-2002 Netfilter core team
Feb 9 14:52:34 www kernel: Intel(R) PRO/100 Network Driver - version
2.3.30-k1
Feb 9 14:52:34 www kernel: Copyright (c) 2003 Intel Corporation
Feb 9 14:52:34 www kernel:
Feb 9 14:52:34 www kernel: PCI: Found IRQ 10 for device 00:03.0
Feb 9 14:52:34 www kernel: e100: eth0: Intel(R) PRO/100 Network Connection
Feb 9 14:52:34 www kernel: Hardware receive checksums enabled
Feb 9 14:52:34 www kernel: cpu cycle saver enabled
Feb 9 14:52:34 www kernel:
Feb 9 14:52:34 www kernel: PCI: Found IRQ 5 for device 00:04.0
Feb 9 14:52:34 www kernel: e100: eth1: Intel(R) PRO/100 Network Connection
Feb 9 14:52:34 www kernel: Hardware receive checksums enabled
Feb 9 14:52:34 www kernel: cpu cycle saver enabled
Feb 9 14:52:34 www kernel:
Feb 9 14:52:34 www kernel: ip_tables: (C) 2000-2002 Netfilter core team
Feb 9 14:52:34 www kernel: Intel(R) PRO/100 Network Driver - version
2.3.30-k1
Feb 9 14:52:34 www kernel: Copyright (c) 2003 Intel Corporation
Feb 9 14:52:34 www kernel:
Feb 9 14:52:34 www kernel: PCI: Found IRQ 10 for device 00:03.0
Feb 9 14:52:34 www kernel: e100: eth0: Intel(R) PRO/100 Network Connection
Feb 9 14:52:34 www kernel: Hardware receive checksums enabled
Feb 9 14:52:34 www kernel: cpu cycle saver enabled
Feb 9 14:52:34 www kernel:
Feb 9 14:52:34 www kernel: PCI: Found IRQ 5 for device 00:04.0
Feb 9 14:52:34 www kernel: e100: eth1: Intel(R) PRO/100 Network Connection
Feb 9 14:52:34 www kernel: Hardware receive checksums enabled
Feb 9 14:52:34 www kernel: cpu cycle saver enabled
Feb 9 14:52:34 www kernel:
Feb 9 14:52:34 www kernel: ip_tables: (C) 2000-2002 Netfilter core team
Feb 9 14:52:34 www kernel: e100: eth0 NIC Link is Up 100 Mbps Full duplex
Feb 9 14:52:34 www kernel: ip_tables: (C) 2000-2002 Netfilter core team
Feb 9 14:52:34 www kernel: e100: eth1 NIC Link is Up 10 Mbps Half duplex
Feb 9 14:52:34 www netfs: Mounting other filesystems: succeeded
Feb 9 14:52:34 www autofs: automount startup succeeded
Feb 9 14:52:35 www insmod:
/lib/modules/2.4.21-9.EL/kernel/drivers/parport/parport_pc.o:
init_module: Device or resource busy
Feb 9 14:52:35 www insmod: Hint: insmod errors can be caused by
incorrect module parameters, including invalid IO or IRQ
parameters. You may find more information in syslog or the output
from dmesg
Feb 9 14:52:35 www insmod:
/lib/modules/2.4.21-9.EL/kernel/drivers/parport/parport_pc.o: insmod
parport_lowlevel failed
--
redhat-list mailing list
unsubscribe mailto:redhat-list-request@(protected)?subject=unsubscribe
https://www.redhat.com/mailman/listinfo/redhat-list
|
|
|