Typically one transaction will require one disk synchronization point and so for real high performance its usually faster to perform many writes within the same transaction to minimize latency caused by disk synchronization. This is especially true when using tools like Berkeley Db or embedded JDBC servers. @param context @throws IOException
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|