On 12/9/06, Sudev Barar <sbarar gmail com> wrote:
On 10/12/06, Bruce Anderson <bruce g anderson gmail com
> I think I did a pretty default installation of K12LTSP version 4.4.1 from
> CDs on a VALinux 1220 (dual 800 MHz PIIIs with 2048MB and two 18 GB scsci
Bruce can you describe your network and confirm you are using 100mbps
switches and not hubs?
Between my experimental LTSP server and my experimental client I have a
direct connect crossover cable no hub, now switch; the message on the
client boot screen indicates 100Mbps full-duplex and /var/log/messages
on the server says the same.
The network may be a problem - I am not really clear how different
networks interact. In my classroom I have a server (not the LTSP
one) that is between the school LAN and my classroom LAN. That
classroom server has 10.6.44.1
on eth0 connected to
the school and 192.168.44.1
connected to a 10/100
switch for 36 computers in my classroom. All the classroom
computers are 192.168.44.x My LTSP server is 192.168.44.57
eth1 by dhcp from the classroom server and is 192.168.0.254
which directly connects to the client.
> I installed an Aopen AON-325 network card from DisklessWorkstations.com on a
> VALinux 420 (700 MHz Celeron).
> It takes my new LTSP server a long time to boot -- there are various
> messages that the BIOS is too old, MPBIOS bug, agpgart: unable to determine
> aperture size, etc., -- it spends a long time at TFTP ..... and says PXE-E32
> TFTP open timeout -- but it finally boots linux and all
BIOS/MPBIOS messages are something you will have to sort out as far as
LTSP is concerned the process starts only on discovery of DHCPD. The
aperture size could be set to 8mb in BIOS if there is any such setting
I fixed the long boot process by changing the BIOS to default boot from
SCSI without first trying the other default BIOS settings. The
former TFTP timeout does not seem relevant.
The time out message is what needs to be looked at from LTSP point of
view as this is indicating slow network. or network conflicts slowing
it down...like both NIC cards of LTSP two cars servers connected to
same physical network. But green indicators are telling us this may
not be so.
My LTSP server still repeatedly
Dec 12 14:36:14 server dhcpd: DHCPDISCOVER from 00:40:f4:e8:e9:24 via eth0
Dec 12 14:36:15 server dhcpd: DHCPOFFER on 192.168.0.253
to 00:40:f4:e8:e9:24 via eth0
while the client is "Searching for server (DHCP)"
It seems strange to me that in the Service Configuration GUI that the
tftp process is checked and says that xinetd is running -- however the
start stop restart etc. buttons in the tool bar are inactive.
K12OSN mailing list
K12OSN redhat com
For more info see <