[Beowulf] Re: What is the right lubricant for computer rack sliding rails?

[David Mathog]

"Robert G. Brown" <rgb at phy.duke.edu> wrote:
> A tough call -- $2400 to $4000 buys one more node, basically, per rack,
> perhaps a 2.5% increase in capacity.  But you still have to spend
> SOMETHING to stack up the nodes one way or another, so you don't really
> save all of this (unless you stack them right up on the floor with no
> rack at all, the bottom node supporting the weight of the entire stack).
> And if you do this and a low node dies, you may well have to shut down
> and disassemble the entire stack to get at the node.

My first cluster, composed of DS10 alphas was like that - sort of. 
There were 10 nodes, but the tallest stack was only 3 units high, and
the stacks sat next to each other on a work bench (3:3:1).  The DS10s
were built like tanks and were very reliable, only one ever needed
service, and that was due to a PS failure early on. For security (theft
and earthquake) they were padlocked to steel cables which ran down
through the access holes and were tied around the heavy metal cable tray.

The uber-pile is a bit of a straw man.  I'm pretty sure that a 40U stack
of (typical) 1U or 2U servers would squish the one(s) on the bottom,
plus it would be very unsafe to work around, especially here in
earthquake country.  At least on our campus empty racks are often (at
least 1x per year) seen in the junk piles, so it wouldn't be hard to
scrounge one.  Any rack is going to be better than a 40U pile!

Regards,

David Mathog
mathog at caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech