Strange result of Group by on real values

2008-01-30 Thread Joris Kinable
Evening, Could someone explain the result of the Query 2? Query 2 does a GROUP BY avgOctets, but still there are duplicate avgOctets in my result set! Boxplot(octets,"AVG",0) is a UDF which returns REAL values. In this case, the real values have no decimals. The only reasonable explanation I could

Re: Query optimization

2008-01-22 Thread Joris Kinable
t>=boxplot.low AND filter.port_dst<=boxplot.low I'll try the 'EXPLAIN' query tonight, but therefor I've got to create a small database first to speed up the results. Thnx in advance for your help. On Jan 22, 2008 8:15 AM, Sebastian Mendel <[EMAIL PROTECTED]> wr

Query optimization

2008-01-21 Thread Joris Kinable
Optimize query I've got one query, which I would like to improve a lot since it takes very long (>24 hours) to execute. Here is the idea: 1. Take the table (other rows in this table are not mentioned for clearity) and remove all duplicate tuple's. This is done by subquery 'filter'. 2. The same qu

creating indexes with myisamchk

2007-11-18 Thread Joris Kinable
Good evening, I've got to create a very large table: 180GB of data has to be stored. In order to to this I'm using the following steps: 1. Create database structure including keys. 2. Disable keys: ALTER TABLE ut_netflow_4 DISABLE KEYS 3. Load data into the database. 4. Generate all index keys: m