ORCL: Case Study Published on Metalink

My only other blog post related to Oracle has been beefed up, cleaned up, and converted into a Case Study (Metalink DocID 369427.1). Actually that happened a couple of weeks ago, but I thought I'd mention here eventually.

I have to thank a couple of people on my team (Hector and Vickie) for pointing out the typos and suggesting improvements and finding some statspack snippets for me.

I think this is the first article on Metalink that I have actually authored myself. Okay, there is also an ancient article from 1991 on Distributed Queries that is likely obsolete and should be archived away (plus I knew even less about what I was talking about back then anyway). It is a boring challenge for me to write serious, technical articles for an external audience, but I have been asked to do more of these and I usually follow orders.

As far as future ORCL related postings go, I may still publish first drafts of potential Case Studies on my blog first to get wider feedback. If I feel safe, I may even disclose non-technical yarns about life as a front-line technical support person working in a fast-paced software company in the early 90's. Hopefully enough has changed since then that I will not expose any proprietary information or individuals even indirectly. I might even write about my theories about the origins of misconceptions or antiquated ideas about an Oracle database such as

- rebuild your indexes frequently
- keep tables and indexes separate for balance IO
- increase enqueue_resources if you see high enqueue waits
- no writes to datafiles that are in hot backup mode
- the shared pool can be too big

But such postings might be rare since I'd like to think about something other than work once in a while.