- Minimize the join operations
- Avoid long join chains
- Avoid joins between big tables
- Reduce data amount as early as possible
- Use data aggregation
- Use data filters where possible
- Avoid huge query result sets
- Data transfer to the UI clients also requires time
- Too many data can be anyway no analyzed properly
- Minimize data transfer between the engines
- Prefer execution of calculation in lower layers, instead of pushing data up and calculate there
- Avoid expensive calculation or data manipulation
- Row based data base expressions are expensive
- Intermediate results needs to be written into buffer and takes additional time
- Complex expressions, e.g IF, CASE statements
Performance Guidelines - General Principles
Subscribe to:
Comments (Atom)
No comments:
Post a Comment