Oracle

A lot of what we do these days is Oracle. Most people who've done Oracle for a while end up building some monitoring and doing some tuning. For an introduction to the kind of tuning we used to do before Statspack came along, see Oracle Database Tuning Statistics. Statspack has to be the foundation of a more up to date approach to Oracle performance, though getting meaningful answers from the thicket of numbers may be difficult. One answer may be to use orca and orca_oracle.

There is also a new section on specific Oracle issues and resolutions.

If you've had a problem mounting an Oracle CD on HP-UX, check here.

Unix

Once upon a time, as the age of the dinosaurs was coming to an end, a breed of Unix machines grew to rule the land. They swept out of their niche environments in Universities across the world and mutated rapidly. They could see how big the dinosaurs were and they strove to follow. For a while, in the short, bright days of the first multi-processor unix kernels these strange species basked in the warmth. From the AT+T heritage came Pyramid, a classic, clean Unix with family members including the Nile and special skills in MPP. From the BSD line came Sequent, flowering into the NUMA-Q. Who could have known, in those happy times, that Sequent would be swallowed by the original dinosaur? Who would have thought that the Pyramid would trip over its DAT drives and fall? Inexorably, the next generation rose, squeezing the gap.

In the early days, my notes were on paper. As I adapted to the new environment, I built new knowledge. As a result, here's most of what I remembered to write down regarding the vendor-specific stuff I learned in the latter days of my Unix career. Alright, HP is hardly "new generation" stuff, and I haven't used it much, but I offer the few things I've been forced to learn about their way of doing things.

And on a different angle, try the jargon file .