I ran findbugs on our code base and it pointed out there are two more Statements that still need to be closed. In this section of the code we run:
preparedStatement = connection.prepareStatement(query);
for 3 different queries, reusing preparedStatement. In the finally block we do close the resource:
finally{
try{
if (resultSet != null)
resultSet.close();
} catch (Exception e) {
exceptionHandler.ignore(e);
}
try {
if (preparedStatement != null)
preparedStatement.close();
} catch(Exception e) {
exceptionHandler.ignore(e);
}
Should the statement be closed before the next connection.prepareStatement(query); or is this findbugs being cautious?
Yes, the statement must be closed before you perform the next connection.prepareStatement. Otherwise, you're losing your reference to the un-closed previous one (aka leaking statements). Wrap a try {} finally {} around each statement use, closing it in the finally.
Related
It is said to be a good habit to close all JDBC resources after usage. But if I have the following code, is it necessary to close the Resultset and the Statement?
Connection conn = null;
PreparedStatement stmt = null;
ResultSet rs = null;
try {
conn = // Retrieve connection
stmt = conn.prepareStatement(// Some SQL);
rs = stmt.executeQuery();
} catch(Exception e) {
// Error Handling
} finally {
try { if (rs != null) rs.close(); } catch (Exception e) {};
try { if (stmt != null) stmt.close(); } catch (Exception e) {};
try { if (conn != null) conn.close(); } catch (Exception e) {};
}
The question is if the closing of the connection does the job or if it leaves some resources in use.
What you have done is perfect and very good practice.
The reason I say its good practice... For example, if for some reason you are using a "primitive" type of database pooling and you call connection.close(), the connection will be returned to the pool and the ResultSet/Statement will never be closed and then you will run into many different new problems!
So you can't always count on connection.close() to clean up.
Java 1.7 makes our lives much easier thanks to the try-with-resources statement.
try (Connection connection = dataSource.getConnection();
Statement statement = connection.createStatement()) {
try (ResultSet resultSet = statement.executeQuery("some query")) {
// Do stuff with the result set.
}
try (ResultSet resultSet = statement.executeQuery("some query")) {
// Do more stuff with the second result set.
}
}
This syntax is quite brief and elegant. And connection will indeed be closed even when the statement couldn't be created.
From the javadocs:
When a Statement object is closed, its
current ResultSet object, if one
exists, is also closed.
However, the javadocs are not very clear on whether the Statement and ResultSet are closed when you close the underlying Connection. They simply state that closing a Connection:
Releases this Connection object's
database and JDBC resources
immediately instead of waiting for
them to be automatically released.
In my opinion, always explicitly close ResultSets, Statements and Connections when you are finished with them as the implementation of close could vary between database drivers.
You can save yourself a lot of boiler-plate code by using methods such as closeQuietly in DBUtils from Apache.
I'm now using Oracle with Java. Here my point of view :
You should close ResultSet and Statement explicitly because Oracle has problems previously with keeping the cursors open even after closing the connection. If you don't close the ResultSet (cursor) it will throw an error like Maximum open cursors exceeded.
I think you may encounter with the same problem with other databases you use.
Here is tutorial Close ResultSet when finished:
Close ResultSet when finished
Close ResultSet object as soon as you finish
working with ResultSet object even
though Statement object closes the
ResultSet object implicitly when it
closes, closing ResultSet explicitly
gives chance to garbage collector to
recollect memory as early as possible
because ResultSet object may occupy
lot of memory depending on query.
ResultSet.close();
If you want more compact code, I suggest using Apache Commons DbUtils. In this case:
Connection conn = null;
PreparedStatement stmt = null;
ResultSet rs = null;
try {
conn = // Retrieve connection
stmt = conn.prepareStatement(// Some SQL);
rs = stmt.executeQuery();
} catch(Exception e) {
// Error Handling
} finally {
DbUtils.closeQuietly(rs);
DbUtils.closeQuietly(stmt);
DbUtils.closeQuietly(conn);
}
No you are not required to close anything BUT the connection. Per JDBC specs closing any higher object will automatically close lower objects. Closing Connection will close any Statements that connection has created. Closing any Statement will close all ResultSets that were created by that Statement. Doesn't matter if Connection is poolable or not. Even poolable connection has to clean before returning to the pool.
Of course you might have long nested loops on the Connection creating lots of statements, then closing them is appropriate. I almost never close ResultSet though, seems excessive when closing Statement or Connection WILL close them.
Doesn't matter if Connection is poolable or not. Even poolable connection has to clean before returning to the pool.
"Clean" usually means closing resultsets & rolling back any pending transactions but not closing the connection. Otherwise pooling looses its sense.
The correct and safe method for close the resources associated with JDBC this (taken from How to Close JDBC Resources Properly – Every Time):
Connection connection = dataSource.getConnection();
try {
Statement statement = connection.createStatement();
try {
ResultSet resultSet = statement.executeQuery("some query");
try {
// Do stuff with the result set.
} finally {
resultSet.close();
}
} finally {
statement.close();
}
} finally {
connection.close();
}
I created the following Method to create reusable One Liner:
public void oneMethodToCloseThemAll(ResultSet resultSet, Statement statement, Connection connection) {
if (resultSet != null) {
try {
if (!resultSet.isClosed()) {
resultSet.close();
}
} catch (SQLException e) {
e.printStackTrace();
}
}
if (statement != null) {
try {
if (!statement.isClosed()) {
statement.close();
}
} catch (SQLException e) {
e.printStackTrace();
}
}
if (connection != null) {
try {
if (!connection.isClosed()) {
connection.close();
}
} catch (SQLException e) {
e.printStackTrace();
}
}
}
I use this Code in a parent Class thats inherited to all my classes that send DB Queries. I can use the Oneliner on all Queries, even if i do not have a resultSet.The Method takes care of closing the ResultSet, Statement, Connection in the correct order. This is what my finally block looks like.
finally {
oneMethodToCloseThemAll(resultSet, preStatement, sqlConnection);
}
With Java 6 form I think is better to check it is closed or not before close (for example if some connection pooler evict the connection in other thread) - for example some network problem - the statement and resultset state can be come closed. (it is not often happens, but I had this problem with Oracle and DBCP). My pattern is for that (in older Java syntax) is:
try {
//...
return resp;
} finally {
if (rs != null && !rs.isClosed()) {
try {
rs.close();
} catch (Exception e2) {
log.warn("Cannot close resultset: " + e2.getMessage());
}
}
if (stmt != null && !stmt.isClosed()) {
try {
stmt.close();
} catch (Exception e2) {
log.warn("Cannot close statement " + e2.getMessage());
}
}
if (con != null && !conn.isClosed()) {
try {
con.close();
} catch (Exception e2) {
log.warn("Cannot close connection: " + e2.getMessage());
}
}
}
In theory it is not 100% perfect because between the the checking the close state and the close itself there is a little room for the change for state. In the worst case you will get a warning in long. - but it is lesser than the possibility of state change in long run queries. We are using this pattern in production with an "avarage" load (150 simultanous user) and we had no problem with it - so never see that warning message.
Some convenience functions:
public static void silentCloseResultSets(Statement st) {
try {
while (!(!st.getMoreResults() && (st.getUpdateCount() == -1))) {}
} catch (SQLException ignore) {}
}
public static void silentCloseResultSets(Statement ...statements) {
for (Statement st: statements) silentCloseResultSets(st);
}
As far as I remember, in the current JDBC, Resultsets and statements implement the AutoCloseable interface. That means they are closed automatically upon being destroyed or going out of scope.
Consider the following code
ResultSet rs = null;
Statement st = null;
try {
//do somehting
} catch (Exception e){
//do something
} finally {
if(st != null){
try {
st.close();
} catch (SQLException e) {
log.error("Exception while closing statement: " + e);
}
}
}
The question is that when we close the statement, will it close the result set as well or do we need to explicitly close the result set like this
if(rs != null){
try {
rs.close();
} catch (SQLException e) {
log.error("Exception while closing result set: " + e);
}
}
I thought that closing the statement will automatically close the result set, but FindBugs throws the following warning if I don't explicitly close the result set
This method may fail to clean up java.sql.ResultSet
When a Statement object is closed, its current ResultSet object, if one exists, is also closed.
This and this shows that Oracle may have problems and you might have to explicitly close the ResultSet. But again, as per the Javadocs, this shouldn't be an issue. Hence, the warning,maybe.
You can't count on the ResultSet being closed automatically, it depends on the driver implementation and how compliant it is. The best policy is to close the ResultSet explicitly.
When you close the statement or connection, all it's children should be closed too by default.
How to integrate the common JDBC idiom of creating/receiving a connection, querying the database and possibly processing the results with Java 7's automatic resource management, the try-with-resources statement? (Tutorial)
Before Java 7, the usual pattern was something like this:
Connection con = null;
PreparedStatement prep = null;
try{
con = getConnection();
prep = prep.prepareStatement("Update ...");
...
con.commit();
}
catch (SQLException e){
con.rollback();
throw e;
}
finally{
if (prep != null)
prep.close();
if (con != null)
con.close();
}
With Java 7 you can go for:
try(Connection con = getConnection(); PreparedStatement prep = con.prepareConnection("Update ..."){
...
con.commit();
}
This will close the Connection and the PreparedStatement, but what about the rollback? I cannot add a catch clause containing the rollback, because the connection is only available within the try block.
Do you still define the connection outside of the try block? What is the best practice here, especially if connection pooling is used?
try(Connection con = getConnection()) {
try (PreparedStatement prep = con.prepareConnection("Update ...")) {
//prep.doSomething();
//...
//etc
con.commit();
} catch (SQLException e) {
//any other actions necessary on failure
con.rollback();
//consider a re-throw, throwing a wrapping exception, etc
}
}
According to the oracle documentation, you can combine a try-with-resources block with a regular try block. IMO, the above example captures the correct logic, which is:
Attempt to close the PreparedStatement if nothing goes wrong
If something goes wrong in the inner block, (no matter what is is) roll back the current transaction
Attempt to close the connection no matter what
If something goes wrong closing the connection, you can't rollback the transaction (as that's a method on the connection, which is now in indeterminate state), so don't try
In java 6 and earlier, I would do this with a triply nested set of try blocks (outer try-finally, middle try-catch, inner try-finally). ARM syntax does make this terser.
IMO, declaring Connection and PreparedStatement outside try-catch is the best way available in this case.
If you want to use pooled connection in transaction, you should use it in this way:
try (Connection conn = source.getConnection()) {
conn.setAutoCommit(false);
SQLException savedException = null;
try {
// Do things with connection in transaction here...
conn.commit();
} catch (SQLException ex) {
savedException = ex;
conn.rollback();
} finally {
conn.setAutoCommit(true);
if(savedException != null) {
throw savedException;
}
}
} catch (SQLException ex1) {
throw new DataManagerException(ex1);
}
This sample code handles setting autocommit values.
NOTE, that using savedException does save exception in case that conn.rollback() throws another. This way, finally block will throw "right" exception.
My application has a memory leak resulting from my usage of JDBC. I have verified this by looking at a visual dump of the heap and seeing thousands of instances of ResultSet and associated objects. My question, then, is how do I appropriately manage resources used by JDBC so they can be garbage collected? Do I need to call ".close()" for every statement that is used? Do I need to call ".close()" on the ResultSets themselves?
How would you free the memory used by the call:
ResultSet rs = connection.createStatement().executeQuery("some sql query");
??
I see that there are other, very similar, questions. Apologies if this is redundant, but either I don't quite follow the answers or they don't seem to apply universally. I am trying to achieve an authoritative answer on how to manage memory when using JDBC.
::EDIT:: Adding some code samples
I have a class that is basically a JDBC helper that I use to simplify database interactions, the main two methods are for executing an insert or update, and for executing select statements.
This one for executing insert or update statements:
public int executeCommand(String sqlCommand) throws SQLException {
if (connection == null || connection.isClosed()) {
sqlConnect();
}
Statement st = connection.createStatement();
int ret = st.executeUpdate(sqlCommand);
st.close();
return ret;
}
And this one for returning ResultSets from a select:
public ResultSet executeSelect(String select) throws SQLException {
if (connection == null || connection.isClosed()) {
sqlConnect();
}
ResultSet rs = connection.createStatement().executeQuery(select);
return rs;
}
After using the executeSelect() method, I always call resultset.getStatement().close()
Examining a heap dump with object allocation tracing on shows statements still being held onto from both of those methods...
You should close the Statement if you are not going to reuse it. It is usually good form to first close the ResultSet as some implementations did not close the ResultSet automatically (even if they should).
If you are repeating the same queries you should probably use a PreparedStatement to reduce parsing overhead. And if you add parameters to your query you really should use PreparedStatement to avoid risk of sql injection.
Yes, ResultSets and Statements should always be closed in a finally block. Using JDBC wrappers such as Spring's JdbcTemplate helps making the code less verbose and close everything for you.
I copied this from a project I have been working on. I am in the process of refactoring it to use Hibernate (from the code it should be clear why!!). Using a ORM tool like Hibernate is one way to resolve your issue. Otherwise, here is the way I used normal DAOs to access the data. There is no memory leak in our code, so this may help as a template. Hope it helps, memory leaks are terrible!
#Override
public List<CampaignsDTO> getCampaign(String key) {
ResultSet resultSet = null;
PreparedStatement statement = null;
try {
statement = connection.prepareStatement(getSQL("CampaignsDAOImpl.getPendingCampaigns"));
statement.setString(1, key);
resultSet = statement.executeQuery();
List<CampaignsDTO> list = new ArrayList<CampaignsDTO>();
while (resultSet.next()) {
list.add(new CampaignsDTO(
resultSet.getTimestamp(resultSet.findColumn("cmp_name")),
...));
}
return list;
} catch (SQLException e) {
logger.fatal(LoggerCodes.DATABASE_ERROR, e);
throw new RuntimeException(e);
} finally {
close(statement);
}
}
The close() method looks like this:
public void close(PreparedStatement statement) {
try {
if (statement != null && !statement.isClosed())
statement.close();
} catch (SQLException e) {
logger.debug(LoggerCodes.TRACE, "Warning! PreparedStatement could not be closed.");
}
}
You should close JDBC statements when you are done. ResultSets should be released when associated statements are closed - but you can do it explicitly if you want.
You need to make sure that you also close all JDBC resources in exception cases.
Use Try-Catch-Finally block - eg:
try {
conn = dataSource.getConnection();
stmt = conn.createStatement();
rs = stmet.executeQuery("select * from sometable");
stmt.close();
conn.close();
} catch (Throwable t) {
// do error handling
} finally {
try {
if (stmt != null) {
stmt.close();
}
if (conn != null) {
conn.close();
}
} catch(Exception e) {
}
}
I have a doubt regarding database operation.I have one insert query that should run for 10 times. the loop starts and inserted 4 or 5 val while inserting 6th, the db connection got failed for a while and again connected. then what will happen,
whether it skips that particular val or throws exception or roll back th entire operation?
EDIT : Sample Code
try
{
String sql_ji_inser="insert into job_input values (?,?)";
PreparedStatement pst_ji_inser=OPConnect.prepareStatement(sql_ji_inser);
for(int i=0;i<v_new_data.size();i++)
{
Vector row=new Vector();
row=(Vector)v_new_data.get(i);
job_id=Integer.parseInt(row.get(0).toString());
item_no=Integer.parseInt(row.get(1).toString());
pst_ji_inser.setInt(1,job_id);
pst_ji_inser.setInt(2,item_no);
pst_ji_inser.addBatch();
}
System.out.println("No of rows inserted"+pst_ji_inser.executeBatch().length);
}
catch(Exception ex)
{
System.out.println("********Insert Exception*********************");
ex.printStackTrace();
return false;
}
Is this the right way
try
{
int count=0;// for checking no of inserting values
OPConnect.setAutoCommit(false);
String sql_ji_inser="insert into job_input values (?,?)";
PreparedStatement pst_ji_inser=OPConnect.prepareStatement(sql_ji_inser);
for(int i=0;i<v_new_data.size();i++)
{
job_id=Integer.parseInt(row.get(0).toString());
item_no=Integer.parseInt(row.get(1).toString());
pst_ji_inser.setInt(1,job_id);
pst_ji_inser.setInt(2,item_no);
pst_ji_inser.addBatch();
count++;
}
int norowinserted=pst_ji_inser.executeBatch().length;
if(count==norowinserted)
{
OPConnect.commit();
}
}
catch(Exception ex)
{
System.out.println("********Insert Exception*********************");
OPConnect.rollback();
ex.printStackTrace();
return false;
}
That depends on how you're inserting the rows. If you're inserting them in a single transaction on a connection which has auto-commit turned off by connection.setAutoCommit(false) and you're commiting the connection after completing the insert queries using connection.commit() and you're explicitly calling connection.rollback() inside the catch block, then the entire transaction will be rolled back. Otherwise, you're dependent on environmental factors you have no control over.
See also:
When to call rollback?
Update: here's a rewrite of your code. Note that the connection and statement should be declared before the try, acquired in the try and closed in the finally. This is to prevent resource leaking in case of exceptions.
String sql = "insert into job_input values (?, ?)";
Connection connection = null;
PreparedStatement statement = null;
try {
connection = database.getConnection();
connection.setAutoCommit(false);
statement = connection.prepareStatement(sql);
for (List row : data) {
statement.setInt(1, Integer.parseInt(row.get(0).toString()));
statement.setInt(2, Integer.parseInt(row.get(1).toString()));
statement.addBatch();
}
statement.executeBatch();
connection.commit();
return true;
} catch (SQLException e) {
if (connection != null) try { connection.rollback(); } catch (SQLException logOrIgnore) {}
e.printStackTrace();
return false;
} finally {
if (statement != null) try { statement.close(); } catch (SQLException logOrIgnore) {}
if (connection != null) try { connection.close(); } catch (SQLException logOrIgnore) {}
}
I am by the way not a fan of returning a boolean here. I'd just make the method void, let the catch throw e and put the calling code in a try-catch.