Big Cluster Node Instructions

If you are the DHCP master, see BigClusterDHCPInstructions.

Please read and follow these instructions!

You've been given this sheet and an index card. The index card is to identify your computer; instructions are later on.

Boot

• Follow directions for trivial-net-setup. Hit Enter to select the highlighted answer and the arrow keys to change the selection.

On the Mac, it will boot you straight through the boot sequence. On x86 machines:

• Boot up to the BCCD splash screen. (When it says boot>, ignore it. After a second, it will continue booting.)
• Hit Enter

Everybody (x86 and PPC both) then:

• Enter the password we give you on the day of the event.
• Follow directions for trivial-net-setup.
  • Say YES/OK (Enter to accept) to everything.
  • When it tells you what your IP address is, write the last bit down in big letters (easy to read from six feet away). So if it tells you that you are 198.162.1.105, write down 105 on the card.
  • Write your name (legibly!) on the index card. If you are not the owner, write the owner's name down also. First name is probably enough.

• When you get to the option of logging in:
  • Login as root, using the password listed at the login prompt
  • Change the password. If you are helping the owner, let the owner set the password.
  • df to get a list of the mounted partitions
  • umount partition for all of your local drive partitions, (e.g. umount /mnt/rw/discs/disc0/part3/home/fred) Macs don't seem to mount any of your local drives.
  • exit
• Sign in as bccd, with the password given earlier.
• Answer yes when it asks if you want to run a heartbeat.
• bccd-allowall (Answer yes.)
• bccd-snarfhosts

Windows users only have the option of

• startx

but Mac users, your trackpad might not work; you might get stuck and hosed. Don't startx.

Compile the target code

Everybody needs to compile the target code.

• Ducky needs to push out the demo code.
• The code will show up in a directory named something like /tmp/6g2w98s.
• cp -r dirname/povray31 ~bccd/ (where dirname is something like /tmp/6g2w98s)

For x86 machines, the code should just run. PPC users (and those who feel inclined to do more) recompile by doing:

• cd ~bccd/povray31/source/libpng
• make -f makefile.lnx clean
• make -f makefile.lnx
• cd ~bccd/povray31/source/mpi-unix
• change the makefile to use the Generic Linux CFLAGS: comment out line 71 and uncomment 65.
• make clean
• make
• cd ~bccd/povray31

• When you are done with all your setup, stick the index card in your keyboard (so that we can easily find a node if we need to).

Other code that we might run:

• hello world
  • cd ~bccd
  • mpicc -o hello hello_*.c NOTE THE UNDERSCORE!
  • mpirun -np 8 -machinefile ~/machines hello
• various code in ~bccd/lam-mpi. Note that it requires a fair amount of gymnastics to switch to LAM.

---

To switch to LAM

To switch to LAM, you need to do this:

• edit ~/.bashrc
  • edit the PATH line so that the line reads export PATH=/lam-mpi/bin:$PATH
  • write file and quit
• source ~/.bashrc (or log out and log back in)
• For each node, rebuild the library cache:
  • su - root (using the root password given)
  • ldconfig -v | less
  • exit (back to bccd)
• bccd-allowall (Answer yes.)
• bccd-snarfhosts
• recon -v ~/machines
• lamboot -v ~/machines

---

As more machines come online, you might need to refresh your system state:

• bccd-allowall
• bccd-snarfhosts
• recon -v ~/machines

There is some order dependency that Ducky hasn't quite figured out yet; keep doing those and eventually it will all get settled out.

Mac users, if you have trouble rebooting into OS X immediately after booting into BCCD, try holding Control-Command-Power after an non-starting boot.