vBulletin Search Engine Optimization
| |||||||
| Register | FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| I am not at all a DB expert but my question is that given that, queries to a DB are properly optimized, DB schema is proper & sufficient hardware resources are in place, how can I make my queries run faster ?? I have queries which count and evaluate statistics from 6 month old customer data (like going through millions of records), sometimes for a 100's of customers and sometimes even fetching all records a month old. The system is MySQL 4.0 based with InnoDB tables. There is no parallelization or clustering concepts used. I know it's a very open-ended question, but in essence how should one speed up the queries. Should I study query optimization algorithms & figure out which commercial/open source DB has the best performing algos & buy that OR study how to write a query to make it illicit a certain algoritm behavior/query plan OR play with DB caching techniques ... What are the steps to be taken ? |
| ||||
| rchadda@gmail.com 寫道: > I am not at all a DB expert but my question is that given that, queries > to a DB are properly optimized, DB schema is proper & sufficient > hardware resources are in place, how can I make my queries run faster > ?? > > I have queries which count and evaluate statistics from 6 month old > customer data (like going through millions of records), sometimes for a > 100's of customers and sometimes even fetching all records a month old. > > The system is MySQL 4.0 based with InnoDB tables. There is no > parallelization or clustering concepts used. > > I know it's a very open-ended question, but in essence how should one > speed up the queries. > > Should I study query optimization algorithms & figure out which > commercial/open source DB has the best performing algos & buy that OR > study how to write a query to make it illicit a certain algoritm > behavior/query plan OR play with DB caching techniques ... What are the > steps to be taken ? the best solution is to cache the queries' results in the memory, e.g. memcached |