best practices cookbook?

[Miles Fidelman]

Hi,

I'm about to rebuild a small cluster (4 servers) that supporting a mix
of development and production, in a tiny organization. The current
installation has evolved over time, with no real structure to it's
operations toolchain.

Over the years, I've played with a lot of monitoring and control tools
(e.g. Nagios, Munin, Webmin, various ISP control panels, etc.) and
I've been doing a lot of reading about new tools in the devops space
(chef, puppet, et. al.).

As I start to sort through what makes sense for refactoring our
cluster, I've found lots of good "intro to devops" slide decks, and
deep dives into various tools, but nothing along the lines of serious
best practices or cookbooks or case studies aimed at small
installations.

Which leads me to wonder, has anybody out there put together, or seen,
a slide deck that presents a bottom up, layer-by-layer walk through of
tools that are available, how to chose, which pieces work together,
and so forth - working up from hardware monitoring, to disk and
filesystem management, to o/s and v/m management, to services, to
applications, to users?

And, I know it's probably way to much to ask, but my installation
starts with Debian and Xen - so anything aimed at that environment
would be really cool.

Thanks for any pointers,

Miles Fidelman