[K12OSN] K12LTSP won't LTSP

Jim McQuillan jam at mcquil.com
Mon Sep 11 11:37:51 UTC 2006


Sonjag,

I see 2 problems:

1) You have eth0 and eth1 both in an overlapping network subnet.

2) And you said that both eth0 and eth0 are plugged into the network.  
I'm assuming you mean that eth0 and eth1 are both plugged into the same 
network.

This is a BAD idea to have 2 nics on the server plugged into the same 
network. Unless, you are doing channel bonding, which it doesn't look 
like you are doing.

When the server wants to send data to one of the workstations, the 
kernel will get confused as to which interface to send the data through.

You really should consider having the server connected to your LAN 
through only 1 interface, and make sure your other interface is either 
disabled, or on a completely separate physical network with a different 
subnet specified.

Jim McQuillan
jam at Ltsp.org



sonjag at comcast.net wrote:
> Thanks to both of you for your ideas. Here is the result of ifconfig:
>
> [root at LTSP ~]# ifconfig
> eth0      Link encap:Ethernet  HWaddr 00:04:23:D3:AA:7E
>           inet addr:172.27.1.30  Bcast:172.27.255.255  Mask:255.255.0.0          inet6 addr: fe80::204:23ff:fed3:aa7e/64 Scope:Link
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>           RX packets:3661004 errors:0 dropped:0 overruns:0 frame:0
>           TX packets:31223 errors:0 dropped:0 overruns:0 carrier:0
>           collisions:0 txqueuelen:100
>           RX bytes:237426619 (226.4 MiB)  TX bytes:2736558 (2.6 MiB)          Base address:0xdc00 Memory:fcfa0000-fcfc0000
>
> eth1      Link encap:Ethernet  HWaddr 00:04:23:D3:AA:7F
>           inet addr:172.27.6.192  Bcast:172.27.0.255  Mask:255.255.0.0          inet6 addr: fe80::204:23ff:fed3:aa7f/64 Scope:Link
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>           RX packets:3679630 errors:0 dropped:0 overruns:0 frame:0
>           TX packets:4381 errors:0 dropped:0 overruns:0 carrier:0
>           collisions:0 txqueuelen:100
>           RX bytes:237973927 (226.9 MiB)  TX bytes:983954 (960.8 KiB)          Base address:0xdc80 Memory:fcfe0000-fd000000
>
> lo        Link encap:Local Loopback
>           inet addr:127.0.0.1  Mask:255.0.0.0
>           inet6 addr: ::1/128 Scope:Host
>           UP LOOPBACK RUNNING  MTU:16436  Metric:1
>           RX packets:16432 errors:0 dropped:0 overruns:0 frame:0
>           TX packets:16432 errors:0 dropped:0 overruns:0 carrier:0
>           collisions:0 txqueuelen:0
>           RX bytes:10439455 (9.9 MiB)  TX bytes:10439455 (9.9 MiB)
>
> eth0 and eth0 are both plugged into my network. I am able to get to the Internet from the server and my terminals do get an IP address, so I believe both ports are properly connected to my network and functioning. 
>
> John, I don't have an iftab in my /etc folder, so I couldn't try your suggestion. I am using K12LTSP, which uses Fedora core 5.
>
> I noticed that they are on different bcast "channels" (not sure if that's the correct term.) Is this alright? Could the LTSP be a problem somewhere else besides the network cards? Any ideas are appreciated.
>
> Thanks--
>
> Sonja
>
> ________________________________________________________
> Huck wrote:
>
>     sonjag comcast net wrote:
>
>         Early warning: I'm a newbie!
>
>
>         I'm trying to set up my K12LTSP server. It's brand new (last year I ran a small number of clients on an old server.) I loaded v5 and all seemed fine until I tried to get clients up and running. The LTSP box is running DHCP, so I know clients can find it, but when it comes to getting the LTSP, nothing's happening.
>
>             From the Fedora doc, I found this:
>
>         By default the LTSP sever runs DHCP on the client's card (eth0) and automatically gives out IP #'s upon request. It then accepts bootp and PXE boot requests and passes on the Linux kernel to the client.
>
>         So I think that means that DHCP and PXE both run on eth0, so the card's working. Maybe not.
>
>         I remember someone (David Trask maybe?) mentioning that you want to have 2 really different ethernet ports or you might have problems. This server has 2 built in Intel gig ethernet ports. Maybe this is the problem. I just don't know how to troubleshoot it. 
>
>     If there are two ethernet ports...
>     try typing this as root as the commandline
>
>     'ifconfig'
>
>     see if it shows settings for eth0 and eth1.
>
>
>     if so...then perhaps you merely need to plug in a network cable from eth0 to the switch your thin clients are plugged into.
>
>     and plug eth1 into your primary network.
>
>
> I believe the more elegant solution to the problem of the system assigning the network device name to the wrong interface (or at least different than it used to) is to specify exactly what you want in /etc/iftab ... (in Debian, anyway)
>
> jhawley steelix:~$ cat /etc/iftab
> # This file assigns persistent names to network interfaces.
> # See iftab(5) for syntax.
>
> eth0 mac 00:40:05:39:01:57 arp 1
> eth1 mac 00:05:5d:d0:82:d7 arp 1
>
>
> ~jh
>
> _______________________________________________
> K12OSN mailing list
> K12OSN at redhat.com
> https://www.redhat.com/mailman/listinfo/k12osn
> For more info see <http://www.k12os.org>
>   




More information about the K12OSN mailing list