Hello all,
I must admit that I am new to the Linux/PXE world, and so, am still learning a
lot. I am currently in the 2nd semester of a senior project at Weber State
University in Utah, trying like mad to get a diskless Beowulf Cluster up and
running. I was given a grant to purchase 4 nodes running this equipment:
Pentium D dual core Processor
MSI 965 Motherboard
Intel Pro/1000 GT PCI adapter
I am using FC6, 64-bit version to boot a diskless linux cluster. Just trying to
get past the problems of booting FC6 on the 965 chipset with and IDE drive was
headache enough, now I am trying to get the DHCP server to connect with the
nodes and pass a kernel and root file system. I have used various versions of
dhcp.conf files to no avail. The Intel Boot Agent version is GE v1.2.22 with a
PXE 2.1 build. I was able to get communication with a different cluster that
was just a P4 cluster with FC5 but got a kernel error on loading that I believe
may be a PXE stack problem because of the 3COM cards in that cluster. Now, when
I attempt to boot a diskless machine on the FC6 cluster, the node will display:
MAC {mac address displayed} GUID:FFFFFFF-FFFF-FFFF-FFFFFFFF
(I have no idea what the GUID means and why I am getting all F's)
DHCP_ _ _ _
Then the node will get its own IP address, the mask, and also the IP address of
the DHCP server and then gives the following error:
PXE-E55 ProxyDHCP server did not reply to request on port 4011.
This is where I have been stopped cold. I have tried different versions of the
DHCP.conf file suggest on the SYSLINUX web page and also downloaded a separate
PXE server to run on port 4011, but still get the same result. I have even
swapped out NIC cards, looked for updates, and tried to find different ways of
passing a kernel and OS but without results. I haven't found very many
clusters using the GE v1.2.22 Intel Boot Agent so this may be a problem.
The dhcp.conf versions I have used are as follows:
The first version that worked with FC5 on the P4 cluster:
ddns-update-style none;
# Definition of PXE-specific options
option space PXE;
option PXE.mtftp-ip code 1 = ip-address;
option PXE.mtftp-cport code 2 = unsigned integer 16;
option PXE.mtftp-sport code 3 = unsigned integer 16;
option PXE.mtftp-tmout code 4 = unsigned integer 8;
option PXE.mtftp-delay code 5 = unsigned integer 8;
option PXE.discovery-control code 6 = unsigned integer 8;
option PXE.discovery-mcast-addr code 7 = ip-address;
subnet 10.10.10.0 netmask 255.255.255.0 {
class "pxeclients" {
match if substring (option vendor-class-identifier, 0, 9) =
"PXEClient";
option vendor-class-identifier "PXEClient";
vendor-option-space PXE;
# At least one of the vendor-specific PXE options must be set in
# order for the client boot ROMs to realize that we are a PXE-compliant
# server. We set the MCAST IP address to 0.0.0.0 to tell the boot ROM
# that we can't provide multicast TFTP (address 0.0.0.0 means no
# address).
option PXE.mtftp-ip 0.0.0.0;
# This is the name of the file the boot ROMs should download.
filename "pxelinux.0";
# This is the name of the server they should get it from.
next-server 10.10.10.1;
}
pool {
max-lease-time 86400;
default-lease-time 86400;
range 10.10.10.2 10.10.10.254;
deny unknown clients;
}
host node1 {
hardware ethernet 00:50:DA:B5:2A:3D;
fixed-address 10.10.10.5;
}
}
I read some prior archives with this error, and it was suggested to use a full
PXE server or try not to use the PXE option space. I then tried to simplify the
dhcpd.conf file to the following:
ddns-update-style none;
allow booting;
allow bootp;
option subnet-mask 255.255.255.0;
group
{
option dhcp-class-identifier "PXEClient";
next-server 10.10.10.1;
filename "pxelinux.0";
host node1
{
hardware ethernet 00:0E:0C:C4:0B:7A;
fixed-address 10.10.10.5;
}
}
But, I still get nothing. Any sort of help would be appreciated as it is
obvious I am new to this and need a little more explanation than others may.
Thanks in advance for the help.
Stephen Jenkins WSU
snj78 at comcast.net