Re: Nearest Common Ancestor Report (XDb1's $1000 Challenge)

From: Neo <>
Date: 29 May 2004 11:18:22 -0700
Message-ID: <>

> There seems to be some puzzling differences in our time measurements.

I believe the reason why measurements made on my machine (500 Mhz) aren't translating into proportionally shorter measurements on your machine (1.3 Ghz) is due to a difference in disk I/O characteristics. XDb1 v4.5.3 (and earlier) writes to the hard drive at the end of each line on the report.

Are the below sequence of events approximately correct when creating the report via SQL Server's Query Analyzer?

  1. Run report procedure
  2. Print Start Time.
  3. Create Report in RAM table.
  4. Issue asynchronous command to save RAM table to HD.
  5. Print End Time.
  6. RAM table continues to be written to HD.
Received on Sat May 29 2004 - 20:18:22 CEST

Original text of this message