Re: Re: Re: [vserver] Networktroubles centos6

From: Roman Pretory <roman.pretory_at_austrianonlines.com>
Date: Sat 17 Mar 2012 - 09:43:38 GMT
Message-ID: <96909ac42385e351a3e575e4f725ed17@10.0.0.6>

Hello Steve

Thanks for you help.
if I have this server back by me, I see that he have troubles with both
networkcards(realtech and intel).
So I make a update of the bios, there was a new 14days ago, also a take the
new kernel,
change the pci card to Intel express what ist should be from the beginn,
and insatlled the intel driver now e1000e from intel.
I deactivate all Hardware not in use in Bios.

After this changes no msg or problem anymore on the server.

The reason that you see only the realtech was that at this time a take the
log only this was inside.

so what one trouble did not come alone, the cisco router on this locations
makes troubles to, it dies without msg,
that meens that by a dhcp leasetime from 4H it could be also a trouble by
geting new Ip by the workstations,
I fixed this by giving fixed ip's and a router change.
The router change was funny because the provider have forgotten or no
remark in his database,
that there is specal config with vlan for the cash machine and send the
wrong tech, without router, soo it thake time... :-)

--
Roman
On Sat, 10 Mar 2012 11:02:40 +1100, Steve Kieu <msh.computing@gmail.com>
wrote:
> Looks like there is resource conflict somewhere
> 
> irq 16: nobody cared (try booting with the "irqpoll" option)
> 
> 
> I did not see the Intel NIC driver in the dmesg - just the real tek one.
My
> guess is that, whether the BIOS there is some sort of  PCI share
resource
> conflict or the kernel driver. It might also be a failing hardware
> component given that this is pretty well tested rehat kernel - the
hardware
> - or BIOS settings is more likely.  Try to boot using irqpoll  kernel
> option to see if it is stable but that case the system will be slower. If
> it is stable then it might be good thing to try newer kernel
> without irqpoll and check dmesg to see if it complains again.
> 
> 
> Hope this help. Let me know if you found something ...
> 
> 
> cheers
> 
> 
> On Sat, Mar 10, 2012 at 9:09 AM, Roman Pretory
> <roman.pretory@austrianonlines.com> wrote:
> This Sever have quit the simples network config from all my Servers
> 
> He have 2 nic's, motherbord default realtech unused and Intel Desktop
> Pro1000GT
> 
> The Intel is changed to be eth0 in the
> /etc/udev/rules.d/70-persistent-net.rules
> 
> eth0 is unconfigured
> br0 have 192.168.1.2/24 GW 192.168.1.1
> Xp-Box  192.168.1.10/24 GW 192.168.1.1 (KVM)
> Samba/VS 192.168.1.11/24 GW 192.168.1.1
> 
> Kernel:kernel 2.6.32-220.4.1.el6.vs2.3.0.36.29.6.21
> Nic driver is loaded as Modul and kernel default // no reload done
> 
> please have a look in dmesg yourself.(dmesg.txt)
> 
> --
> Roman
> 
> 
> 
> 
> 
> On Sat, 10 Mar 2012 08:36:31 +1100, Steve Kieu <msh.computing@gmail.com>
> wrote:
> 
>> What is the network configuration for the window XP box? How the network
> in
>> the host machine configured like how many NICs, and IP assignment to
>> vserver/ the kvm etc..
>>
>> Instead of rebooting if your NIC driver is loaded as module, did you try
> to
>> unload/reload it and see if problem  disappear?
>>
>>
>> I got some network problems before with stock kernel 2.6.32.54 something
>> (from kernel.org and with vserver patch) in some cases - esp. multiple
>> nics/networks gateway setup. I swtich to 2.6.37.5 and no problem so far.
>>
>>
>> Did you see anything in the dmesg log? Something abt warn slow path in
>> networking stack?
>>
>>
>>
>>
>>
>>
>> On Fri, Mar 9, 2012 at 11:43 PM, Roman Pretory
>> <roman.pretory@austrianonlines.com> wrote:
>> Hello
>>
>> I have in the moment curious networktroubles
>>
>> The system is a Centos 6.2 fully updated.
>>
>> Kernel and Utils are from rpm.hozac.com
>>
>> on this server is running a Vserver Centos 5 with samba und a qemu with
>> windows xp
>>
>> It's running without any troubles 1 Month or so now,
>> normaly I have 60Mbit by samba file copy.
>>
>> if the server is resarted it begin after 2 hourers or a days later that
>> they have errors in the connection.
>> in this time I have only 128kbit by file copy and the ping to the server
>> is
>> on gbLan over 90ms.
>>
>> The strange is that the virtuell xp have no troubles, all is running
> fine,
>> if I stop all virtual servers the ping to the hostsystem makes no
> change,
>> only reboot bring back normal status.
>>
>> I have looked in top, netstat, logs nothing, also stopping alles
>> unnesessary service make no change..
>>
>> so I have change the networkcard from an intel Destopadapter to a
>> realtech,
>> the result is that the troubles came faster back.
>>
>> The last idea I have is that the motherboard is overclocking the pci
> bus..
>> and i have to change to a pci-e card
>>
>> if anyone have an better idea for me I would be happy.
>>
>> --
>> Roman
>>
>>
>>
>>
>>
>>
>>
>>
>>
> 
> 
> 
> 
> 
> 
>
Received on Sat Mar 17 09:43:48 2012
[Next/Previous Months] [Main vserver Project Homepage] [Howto Subscribe/Unsubscribe] [Paul Sladen's vserver stuff]
Generated on Sat 17 Mar 2012 - 09:43:48 GMT by hypermail 2.1.8