As a database-based software architecture, it is inevitable that you must pay attention to the processing and reading of data during design. If the processing is not good, this is often one of the key to the system running performance bottleneck. As a database management and maintenance personnel, it is necessary to obtain the cause of the database to be analyzed by the database, which is generally used as a reference, but this is not the only way, I think it is simply The design architecture of the system, as well as when writing a program, is related to attitude. Database performance analysis: 1. Data buffer cache indicator: Buffer hits, V $ Sysstat Select Round ((1- (SUM (Decode (Name, 'Physical Reads', Value, 0)) / (SUM (Decode Name, 'DB Block gets', value, 0)) SUM (Decode (Name, 'Consistent Gets', Value, 0)))))))))))))))))))))))))))))))))))))))))))))))), 2) "Data Buffer Hit Ratio" from v $ sysstat; = 0.77 Not good, close to 1,> 0.95. Metric Oracle's relative number of data blocks in memory. Improved: Increase DB_BLOCK_BUFFERS 13: Database Initialization Parameters: Buffers = 40960, size = 8192K, a total of about 300M. 2. Reservoir cache is used to cache SQL, V $ librarycache metrics: Pins / Reloads / hit rate Select Sum (Pins "Total Pins", SUM (Reloads) "Total Reloads", Round ((1-SUM (Reloads) / SUM (PINS), 2) "Hit Ratio" from V $ librarycache = 0.99
Select SUM (Pins) / (SUM (Pins) SUM (Reloads)) from V $ librarycache performance is better
