druid連接池獲取不到連接的一種情況

數據源一開始配置：

jdbc.initialSize=1
jdbc.minIdle=1
jdbc.maxActive=5

程序運行一段時間后，執行查詢拋如下異常：

exception=org.mybatis.spring.MyBatisSystemException: nested exception is org.apache.ibatis.exceptions.PersistenceException:
### Error querying database. Cause: org.springframework.jdbc.CannotGetJdbcConnectionException: Could not get JDBC Connection; nested exception is com.alibaba.druid.pool.GetConnectionTimeoutException: wait millis 60000, active 5, maxActive 5
### The error may exist in ...

懷疑是最大連接數不夠，講配置改為

jdbc.initialSize=1
jdbc.minIdle=1
jdbc.maxActive=20

運行一段時間后，又出現上述異常。在datasource配置中添加如下參數（參考http://www.cnblogs.com/netcorner/p/4380949.html）：

<!-- 超過時間限制是否回收 --> 
<property name="removeAbandoned" value="true" /> 
<!-- 超時時間；單位為秒。180秒=3分鍾 --> 
<property name="removeAbandonedTimeout" value="180" /> 
<!-- 關閉abanded連接時輸出錯誤日志 --> 
<property name="logAbandoned" value="true" />

拋如下異常：

2016-12-27 14:35:22.773 [Druid-ConnectionPool-Destroy-1821010113] ERROR com.alibaba.druid.pool.DruidDataSource - abandon connection, owner thread: quartzScheduler_Worker-1, connected time nano: 506518587214834, open stackTrace
at java.lang.Thread.getStackTrace(Thread.java:1552)
at com.alibaba.druid.pool.DruidDataSource.getConnectionDirect(DruidDataSource.java:1014)
at com.alibaba.druid.filter.FilterChainImpl.dataSource_connect(FilterChainImpl.java:4544)
at com.alibaba.druid.filter.stat.StatFilter.dataSource_getConnection(StatFilter.java:662)
at com.alibaba.druid.filter.FilterChainImpl.dataSource_connect(FilterChainImpl.java:4540)
at com.alibaba.druid.pool.DruidDataSource.getConnection(DruidDataSource.java:938)
at com.alibaba.druid.pool.DruidDataSource.getConnection(DruidDataSource.java:930)
at com.alibaba.druid.pool.DruidDataSource.getConnection(DruidDataSource.java:102)
at com.xxx.doCopyIn(AppQustionSync.java:168)
at com.xxx.executeInternal(AppQustionSync.java:99)
at org.springframework.scheduling.quartz.QuartzJobBean.execute(QuartzJobBean.java:75)
at org.quartz.core.JobRunShell.run(JobRunShell.java:202)
at org.quartz.simpl.SimpleThreadPool$WorkerThread.run(SimpleThreadPool.java:573)

紅色標注處連接未釋放，查看代碼，

    try {
                    InputStream input = new FileInputStream(dataFile);
                    conn = copyInDataSource.getConnection();
                    baseConn = (BaseConnection) conn.getMetaData()
                            .getConnection();
                    baseConn.setAutoCommit(false);
                    stmt = baseConn.createStatement();
                    LOG.info("delete data: " + delSql);
                    stmt.executeUpdate(delSql);
                    CopyManager copyManager = new CopyManager(baseConn);
                    LOG.info("copy in: " + copyInSql);
                    copyManager.copyIn(copyInSql, input);
                    baseConn.commit();
                    jobDataMap.remove(data_file_path_key);
                } catch (SQLException e) {
                    try {
                        LOG.warn(JobDataMapHelper.jobName(jobDataMap) + ":"
                                + "批量更新任務失敗回滾...");
                        baseConn.rollback();
                    } catch (SQLException ex) {
                        String errorMessage = String.format(
                                "JobName:[%s] failed: ",
                                JobDataMapHelper.jobName(jobDataMap));
                        throw new SQLException(errorMessage, ex);
                    }
                } finally {
                    stmt.close();
 baseConn.close();  
                    jobDataMap.remove(data_file_path_key);
                    FileUtils.deleteQuietly(dataFile);
                }

這里baseConn是通過druid datasource獲取的一個PgSQL 底層連接， 上面代碼執行完后，finally中調用baseConn.close()關閉了這個連接，（猜測關閉這個底層連接，druid連接池卻不知道，還認為自己擁有這個連接，但實際該連接是不可用的，這段代碼執行多次，就將druid連接池中所有連接都耗光了，於是便拋出Could not get JDBC Connection錯誤。

將上面finally代碼塊改為

finally {
                    stmt.close();
                    conn.close();  
                    jobDataMap.remove(data_file_path_key);
                    FileUtils.deleteQuietly(dataFile);
                }

即只調用druid 連接的close方法（只是釋放該連接，而不是直接關閉底層連接），問題解決。