USING AN SQL WORKLOAD - HTTP Programming Recipes for Java Bots - page 509

The releaseStatement method first obtains a mutex . This is the same mutex
used by the obtainStatement function. This ensures that no two threads are access-
ing the PreparedStatement cache at the same time.
try {
try {
this.mutex.acquire();
} catch (InterruptedException e) {
return;
}
The PreparedStatement is added to the cache.
this.statementCache.add(stmt);
Finally, the mutex is released.
} finally {
this.mutex.release();
}
Once the releaseStatement method is complete, the PreparedStatement
is back on the cache awaiting the next thread that needs it.
Implementing an SQL Based Workload Manager
The SQL based workload manager is implemented in the SQLWorkloadManager
class. In this section, I will show you some of the internals of how this class works. This
class makes extensive use of SQL. Teaching extensive SQL is beyond the scope of this topic;
however, many of the methods and functions of the SQLWorkloadManager are simply
wrappers for SQL statements. All of the SQL statements used by the spider are relatively
simple, if you know SQL basics. This section will focus on explaining how the class works and
not what the actual SQL does.
The SQL queries used by this class should work with nearly any SQL database, once the
database is set up. For more information about how to set up the database, see Appendix F.
The SQLWorkloadManager class is shown in Listing 15.2.
Listing 15.2: SQL Workload Management (SQLWorkloadManager.java)
package com.heatonresearch.httprecipes.spider.workload.sql;
import java.net.*;
import java.sql.*;
import java.util.*;
import java.util.concurrent.*;
import java.util.logging.*;
import com.heatonresearch.httprecipes.spider.*;