<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 5.50.4522.1800" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV>
<DIV><FONT face=Arial size=2>People,</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>The start poin is:</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>I´m new in beowulf and i´ve decided to construct a
beowulf machine based in the tutorial of Starling (see </FONT><A
href="http://www.cacr.caltech.edu/beowulf"><FONT face=Arial
size=2>http://www.cacr.caltech.edu/beowulf</FONT></A><FONT face=Arial size=2>
):</FONT></DIV>
<DIV><FONT face=Arial size=2>The tutorial is divided in:</FONT></DIV>
<DIV><FONT face=Arial size=2>1) Introduction of Beowulf:</FONT></DIV>
<DIV><FONT face=Arial size=2>2) Planning the system: About the hardware and some
informations about the partition configuration, etc...</FONT></DIV>
<DIV><FONT face=Arial size=2>3) Software Installation: </FONT></DIV>
<DIV><FONT face=Arial size=2>It´s the maily point of my doubts. </FONT></DIV>
<DIV><FONT face=Arial size=2>I´ve cut and paster the item 3 of the tutorial (see
bellow):</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial
size=2>/-------------------------------------------</FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>To make the system as easy as
possible to maintain, plan on making all compute nodes identical, i.e. same
partition sizes, same software and same configuration files. This makes the
software a whole lot easier to install and maintain. For partition sizes, I like
to give 128 MB to root, 2X memory for swap partition for less than 128 MB RAM,
1X memory if you have 128 MB or more RAM. Some people would say, "Why have any
swap space at all?" For most usage, swapping should not happen on the compute
nodes, but accidents happen. For /usr partition it really depends on the number
of packages you are planning to install, about 800 MB should go a long way for
most systems.</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2>I recommend installing at least the
following, if you are installing </FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2>RedHat or Mandrake
Linux:</FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2> Basic System Software,
including networking software <BR> Kernel Sources <BR> C, C++, g77
compilers and libraries <BR> X11 with development libraries <BR> xntp
or another time synchronizer <BR> autofs <BR> rsync </FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2>To install these packages, please
follow the standard instructions that come with the Linux distribution you
choose to use. You want to try to install all you need, before you start the
cloning process. There are a few files that you probably want to edit before
cloning the system:</FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2> /etc/hosts
<BR> /etc/hosts.equiv <BR> /etc/shosts.equiv <BR> /etc/fstab
<BR> /etc/pam.d/rlogin <BR> /etc/ntp.conf <BR> /etc/lilo.conf
<BR> /etc/exports <BR> /etc/auto.master <BR> /etc/auto.beowulf
<BR></FONT></DIV>
<DIV><FONT face=Arial
size=2>-------------------------------------------/</FONT></DIV>
<DIV><FONT face=Arial size=2>I´ve install the first machine. I labed it of
node00. The clones will be named node01, node02, ..., node07. (My Beowulf will
be 8 PC´s)</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Next step:</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>For the hosts file, you need to
list all the names, with aliases, for each node. All nodes need to be listed in
hosts.equiv and shosts.equiv to let users use rsh commands without creating
their own .rhosts and .shosts files. In your /etc/pam.d/rlogin file, you want to
make sure you do not require secure tty when logging into the nodes. To fix
this, I just switch the order of the modules pam_securetty.so and
pam_rhosts_auth.so., and end up with the following:</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>auth sufficient
/lib/security/pam_rhosts_auth.so <BR>auth required
/lib/security/pam_securetty.so <BR>auth required /lib/security/pam_pwdb.so
shadow nullok <BR>auth required /lib/security/pam_nologin.so <BR>account
required /lib/security/pam_pwdb.so <BR>password required
/lib/security/pam_cracklib.so <BR>password required /lib/security/pam_pwdb.so
shadow nullok use_authtok <BR>session required /lib/security/pam_pwdb.so
</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>When setting up ntp, have the
front-end machine (could be node 0) sync up to the outside world and broadcast
the time to the private network. The nodes will be configured to sync their time
to the front-end machine. If you don't use a private network, you can have all
machines sync up to any ntp server.</EM></FONT></DIV>
<DIV><FONT color=#ff0000></FONT> </DIV>
<DIV><FONT face=Arial
size=2>/-------------------------------------------------</FONT></DIV>
<DIV><FONT face=Arial size=2>I´ve done this...</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Next step:</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>We decided to export scratch
partitions from all nodes and use autofs to mount these from any node upon
request. Our exports file looks like:</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>/scratch
n???.cacr.caltech.edu(rw,no_root_squash)</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>The no_root_squash option lets
root on one node operate as root on any of the mounted files
systems.</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>To set up autofs, add a line
like:</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000
size=2><EM>/data
/etc/auto.beowulf --timeout
600</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>to the file /etc/auto.master and
create /etc/auto.beowulf with entries like:</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>n000 -fstype=nfs
n000:/scratch<BR>n001 -fstype=nfs n001:/scratch<BR>n002 -fstype=nfs
n002:/scratch<BR>n003 -fstype=nfs n003:/scratch</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>On the nodes, you will also want
to automatically mount home directories from the front-end machine, so you'll
need to have:</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>home -fstype=nfs
front-end-machine-name:/home</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>added to the /etc/auto.beowulf
file as well.</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>On the front-end machine I create
a directory named "local" on the large partition used for home directories, move
everything from /usr/local to this directory and make a symbolic link to this
directory from the /usr directory (ex. ln -s /home/local /usr/local). For
software packages like MPICH and PVM, I install them under /usr/local/mpich and
/usr/local/pvm on the front-end machine On the nodes /usr/local will will be a
symbolic link to the NFS mounted file system on the front-end machine. This way
I only have to maintain the packages in one place. PVM can be found at
</EM></FONT><A href="http://www.epm.ornl.gov/pvm"><FONT face=Arial color=#ff0000
size=2><EM>http://www.epm.ornl.gov/pvm</EM></FONT></A><FONT face=Arial
color=#ff0000 size=2><EM> and MPICH can be found at </EM></FONT><A
href="http://www.mcs.anl.gov/mpi/mpich"><FONT face=Arial color=#ff0000
size=2><EM>http://www.mcs.anl.gov/mpi/mpich</EM></FONT></A><FONT face=Arial
color=#ff0000 size=2><EM>.</EM></FONT></DIV>
<DIV><FONT color=#ff0000></FONT> </DIV>
<DIV><FONT face=Arial size=2>/------------------------</FONT></DIV>
<DIV><FONT face=Arial size=2>My partition configuration is:</FONT></DIV>
<DIV><FONT face=Arial size=2>HD total: 6GB</FONT></DIV>
<DIV><FONT face=Arial size=2>Swap: 400MG</FONT></DIV>
<DIV><FONT face=Arial size=2>/root 1.5
GB</FONT></DIV>
<DIV><FONT face=Arial
size=2>/usr 1GB</FONT></DIV>
<DIV><FONT face=Arial size=2>/home 3
GB</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>After configure the files as described before, i
istall the first clone a named it node01 as described in item 4 of the
tutorial:</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Item 4 of the tutorial:</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>4. The Cloning Process</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>There are different ways to clone
your first system. One way is to physically connect a drive to an already
configured system and then copy the content from one drive to another. Another,
maybe more elegant method, is to boot the nodes like diskless clients the first
time and let setup scripts partition and copy the system from tar files or
another already running system.</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>The first method is simpler, but
it requires a lot of reboots and power shut downs. This is the only way to do it
if you don't have a floppy drive in every system. When copying from one drive to
another there are two simple ways of doing this:</EM></FONT></DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM></EM></FONT> </DIV>
<DIV><FONT face=Arial color=#ff0000 size=2><EM>1. Use dd on the entire drive. If
your configured drive is hda and the unconfigured is hdb, you would do "dd
if=/dev/hda of =/dev/hdb". This will take care of partitions as well as boot
sector and files.</EM></FONT></DIV>
<DIV><FONT color=#ff0000></FONT> </DIV>
<DIV><FONT face=Arial size=2>I used the command dd to clone the
system.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Testing the systems:</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>After doing everything described i expted my clone
run well. I´ve tested criating a user named teste in node00 and node 01. When i
log in node01 i criate a directory named sun (/home/geraldo/sun). As
described before the sun wuld be created in the node00, wuld not? I´m doing
something wrong... The exports isn´t running... </FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>It´s because my question about scratch partition,
etc...Then, now i´m studing the linux configuration, export, autofs,
etc.</FONT></DIV>
<DIV><FONT face=Arial size=2>I think after understand the nfs, int teory an
practice, i´ll be able to complete my beowulf system...</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>So, any hint will be great...mainly of someone that
construt the beowul based in the tutorial of Starling....</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>I´m using the online tutorial of linux (Linux
Howto, Deja News, etc), but some points´re dificult to be
solved...</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Obs: I´m sorry my poor english...</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Thanks,</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Geraldo Pereira de Souza (</FONT><A
href="mailto:geraldo@cic.unb.br"><FONT face=Arial
size=2>geraldo@cic.unb.br</FONT></A><FONT face=Arial size=2>)<BR>Laboratório de
Sistemas Integrados e Concorrentes - LAICO<BR>Universidade de Brasília -
UNB</FONT></DIV>
<DIV><FONT face=Arial size=2>Brazil.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>----- Original Message ----- </FONT>
<DIV><FONT face=Arial size=2>From: "Mark Hahn" <</FONT><A
href="mailto:hahn@physics.mcmaster.ca"><FONT face=Arial
size=2>hahn@physics.mcmaster.ca</FONT></A><FONT face=Arial
size=2>></FONT></DIV>
<DIV><FONT face=Arial size=2>To: "Carlos O'Donell Jr." <</FONT><A
href="mailto:carlos@baldric.uwo.ca"><FONT face=Arial
size=2>carlos@baldric.uwo.ca</FONT></A><FONT face=Arial size=2>></FONT></DIV>
<DIV><FONT face=Arial size=2>Cc: <</FONT><A
href="mailto:beowulf@beowulf.org"><FONT face=Arial
size=2>beowulf@beowulf.org</FONT></A><FONT face=Arial size=2>></FONT></DIV>
<DIV><FONT face=Arial size=2>Sent: Monday, December 10, 2001 6:17
PM</FONT></DIV>
<DIV><FONT face=Arial size=2>Subject: Re: Scratch
partition...</FONT></DIV></DIV>
<DIV><FONT size=2><BR><FONT face=Arial></FONT></FONT></DIV><FONT face=Arial
size=2>> > mount -t tmpfs /dev/shm /tmp<BR>> > <BR>> > And in
theory, the numbers I've seen tend to show that this is a bit<BR>> >
faster than having /tmp in the vfs cache path. Though I'm not an <BR>> >
expert on the topic.<BR>> <BR>> generally, filesystems try to be
synchronous with metadata,<BR>> that is, directory entries, inodes,
etc. so you'd be avoiding<BR>> that traffic by using tmpfs. file
*contents* tend to be lazily<BR>> written out by filesystems anyway, so
there's probably <BR>> no savings there. (depends on timing)<BR>>
<BR>> > However, I have noticed that on our compile box<BR>> >
(since gcc tends to toss it's temp .o files in /tmp) it's a bit lighter<BR>>
> on the drive.<BR>> <BR>> gcc -pipe avoids that traffic...<BR>>
<BR>> _______________________________________________<BR>> Beowulf mailing
list, </FONT><A href="mailto:Beowulf@beowulf.org"><FONT face=Arial
size=2>Beowulf@beowulf.org</FONT></A><BR><FONT face=Arial size=2>> To change
your subscription (digest mode or unsubscribe) visit </FONT><A
href="http://www.beowulf.org/mailman/listinfo/beowulf"><FONT face=Arial
size=2>http://www.beowulf.org/mailman/listinfo/beowulf</FONT></A></DIV></BODY></HTML>