Setting up the Master Node

The Box cluster machine does not have a CD-ROM drive. One option is to attach a CD/DVD drive to the SATA slot. The other option is to boot this node through netboot (PXE). To do this, we should use another machine. In our case, we use the first option - viz. adding a CD-ROM drive.

After this, we setup a) the DHCP server and b) the NFS server.

We need to do some network setup for IP masquerading

DHCP server

A DHCP server is needed to dynamically assign IP addresses to the slave-nodes during the PXE boot process. We install it using

apt-get install dhcp3-server

We use the following configuration in /etc/dhcp/dhcpd.conf

ddns-update-style none;
default-lease-time 600;
max-lease-time 7200;
authoritative;
ddns-update-style none;
subnet 192.168.1.0 netmask 255.255.255.0 {
     default-lease-time 86400;
     range 192.168.1.2 192.168.1.20;
     max-lease-time 604800;
     option routers 192.168.1.1;
     option subnet-mask 255.255.255.0;
     option domain-name "local";
     option domain-name-servers 163.143.1.100;
     option nis-domain "hpc";
     option broadcast-address 192.168.1.255;
     allow booting;
     allow bootp;
     if (substring (option vendor-class-identifier, 0, 20)
                     = "PXEClient:Arch:00002") {
             # ia64
             filename "elilo.efi";
             next-server 192.168.1.1;
     } elsif ((substring (option vendor-class-identifier, 0, 9)
                     = "PXEClient") or
             (substring (option vendor-class-identifier, 0, 9)
                     = "Etherboot")) {
             # i386 and x86_64
             filename "pxelinux.0";
             next-server 192.168.1.1;
     } else {
             filename "/install/sbin/kickstart.cgi";
             next-server 192.168.1.1;
     }

    host hpcs02 {
        hardware ethernet 00:15:17:31:11:00;
        option host-name "hpcs02";
        fixed-address 192.168.1.2;
    }
    host hpcs03 {
        hardware ethernet 00:15:17:31:0D:8C;
        option host-name "hpcs03";
        fixed-address 192.168.1.3;
    }
# and so on .........
}

We should edit /etc/network/interfaces and add the following

auto eth1
iface eth1 inet static
     address 192.168.1.1
     netmask 255.255.255.0
     network 192.168.1.0
     broadcast 192.168.1.255
     gateway 163.143.166.104
    dns-nameservers 163.143.1.100

Now restart the network and the dhcp server

service networking restart
service isc-dhcp-server restart

More network configuration

Edit /etc/hosts to include all the nodes. For example

127.0.0.1	localhost
192.168.1.1  hpcs01.u-aizu.ac.jp hpcs01
192.168.1.2  hpcs02.u-aizu.ac.jp hpcs02
192.168.1.3  hpcs03.u-aizu.ac.jp hpcs03
192.168.1.4  hpcs04.u-aizu.ac.jp hpcs04
192.168.1.5  hpcs05.u-aizu.ac.jp hpcs05
192.168.1.6  hpcs06.u-aizu.ac.jp hpcs06
192.168.1.7  hpcs07.u-aizu.ac.jp hpcs07

Also create a text file in /etc/machines with the names of all the slave nodes for example. We will use this for scripting common tasks across the nodes.

hpcs01
hpcs02
...etc

NFS server

Install the packages: apt-get nfs-common nfs-kernel-server

Edit /etc/exports to export /home and /var/lib/tftpboot:

/home     192.168.1.0/24(rw,no_root_squash,sync,no_subtree_check)
/var/lib/tftpboot 192.168.1.0/24(rw,no_root_squash,sync,no_subtree_check)

Export these files with exportfs -av

IP Masquerading

First edit /etc/sysctl.conf and uncomment the following line net.ipv4.ip_forward=1

Follow this by sudo sysctl -p

For IP masquerading we use sudo iptables -t nat -A POSTROUTING -s 192.168.1.0/16 -p eth0 -j MASQUERADE

iptables-save > /etc/network/iptables

For permanent IP masquerading, append the following to /etc/network/interfaces as follows

auto eth1
iface eth1 inet static
address 192.168.0.1
netmask 255.255.255.0
network 192.168.0.0
broadcast 192.168.0.255
up iptables-restore < /etc/network/iptables