Comments on: 42

By: Plans gone AWRy – an invASHtigation « OraStory

Plans gone AWRy – an invASHtigation « OraStory — Tue, 29 Dec 2009 16:37:42 +0000

[...] some organisations, by the time you’ve got access five or six days might have passed. 35 and 42 days are oft-mentioned sensible suggestions, mostly so that a) you’ve got a decent time [...]

By: Jonathan Lewis

Jonathan Lewis — Sat, 01 Mar 2008 14:31:15 +0000

Marco, There's alway an argument for keeping everything "just in case". If you can, and your queries, and the runtime for taking snapshots and running the spreport, don't start taking up excess resources you can just do it. One strategy that I've seen from time to time is the hourly snapshot for a few weeks, then a daily snapshot for a year or so. You don't have the detail, of course, but for slow trends it can be helpful. 9i makes it easy to implement this strategy as it allows you to label a snapshot as a "baseline" - and the sppurge.sql script doesn't delete snapshots that are baselines. (Don't forget to coalesce indexes from time to time, though - and you may want to see if you can do something sensible with the setting for pctused on the tables).

By: Robert Shepard

Robert Shepard — Fri, 29 Feb 2008 19:06:07 +0000

Later this year, the U.S. Postal Service will be charging 42 cents for a first-class stamp. I’m seriously hoping they’ll adopt a Hitchhiker’s Guide theme.

Jonathan: I loved your RMOUG presentations. I’ve been playing with optimizer statistics all week, just trying to wrap my brain around the concept.

There are 10 kinds of people in this world: those who understand binary, and those who don’t.

101010

By: Mark W. Farnham

Mark W. Farnham — Fri, 29 Feb 2008 12:36:58 +0000

Thanks for the base 13 reference. Now I won’t be 42 until December, while I’ve been thinking I was already over 50!

By: Marco Gralike

Marco Gralike — Fri, 29 Feb 2008 01:29:10 +0000

“Grumpy” aka Marco Gralike ;-)

By: Marco Gralike

Marco Gralike — Fri, 29 Feb 2008 01:28:13 +0000

I am 42 and a “grumpy old men”. But what would give me a year’s data (same period a year ago) when the actual date of the changing trending profile happened 159 days ago I didn’t see it on time and / or the business noticed (unforeseeable) / “detected” it for me, in that time period. Business demands follow other rules then actual fact.

By: Noons

Noons — Fri, 29 Feb 2008 01:11:03 +0000

Found it useful to use the statspack export facility: that way I can export and clean up every six months and import the data back to my little play database in the laptop, where I can analyze it to my heart’s content.

By: Flado

Flado — Thu, 28 Feb 2008 22:39:29 +0000

I’m keeping six months of Statspack data online, and I never needed anything more than gathering Optimizer statistics to make sure my queries return in reasonable time (seconds). Of course, I’m always careful to do complete joins, including DBID and INSTANCE_NUMBER even on this non-RAC database. I find the ‘natural join’ syntax very helpful for this.
Thanks for the Douglas Adams reference – it’s always reassuring to know this great man’s words live on.