I'm having a problem closing my database before an attempt to delete the file. The code is just
myconnection.Close();
File.Delete(filename);
And the Delete throws an exception that the file is still in use. I've re-tried the Delete() in the debugger after a few minutes, so it's not a timing issue.
I have transaction code but it doesn't run at all before the Close() call. So I'm fairly sure it's not an open transaction. The sql commands between open and close are just selects.
ProcMon shows my program and my antivirus looking at the database file. It does not show my program releasing the db file after the close().
Visual Studio 2010, C#, System.Data.SQLite version 1.0.77.0, Win7
I saw a two year old bug just like this but the changelog says it's fixed.
Is there anything else I can check? Is there a way to get a list of any open commands or transactions?
New, working code:
db.Close();
GC.Collect(); // yes, really release the db
bool worked = false;
int tries = 1;
while ((tries < 4) && (!worked))
{
try
{
Thread.Sleep(tries * 100);
File.Delete(filename);
worked = true;
}
catch (IOException e) // delete only throws this on locking
{
tries++;
}
}
if (!worked)
throw new IOException("Unable to close file" + filename);
Encountered the same problem a while ago while writing a DB abstraction layer for C# and I never actually got around to finding out what the issue was. I just ended up throwing an exception when you attempted to delete a SQLite DB using my library.
Anyway, this afternoon I was looking through it all again and figured I would try and find out why it was doing that once and for all, so here is what I've found so far.
What happens when you call SQLiteConnection.Close() is that (along with a number of checks and other things) the SQLiteConnectionHandle that points to the SQLite database instance is disposed. This is done through a call to SQLiteConnectionHandle.Dispose(), however this doesn't actually release the pointer until the CLR's Garbage Collector performs some garbage collection. Since SQLiteConnectionHandle overrides the CriticalHandle.ReleaseHandle() function to call sqlite3_close_interop() (through another function) this does not close the database.
From my point of view this is a very bad way to do things since the programmer is not actually certain when the database gets closed, but that is the way it has been done so I guess we have to live with it for now, or commit a few changes to System.Data.SQLite. Any volunteers are welcome to do so, unfortunately I am out of time to do so before next year.
TL;DR
The solution is to force a GC after your call to SQLiteConnection.Close() and before your call to File.Delete().
Here is the sample code:
string filename = "testFile.db";
SQLiteConnection connection = new SQLiteConnection("Data Source=" + filename + ";Version=3;");
connection.Close();
GC.Collect();
GC.WaitForPendingFinalizers();
File.Delete(filename);
Good luck with it, and I hope it helps
Just GC.Collect() didn't work for me.
I had to add GC.WaitForPendingFinalizers() after GC.Collect() in order to proceed with the file deletion.
Had a similar issue, though the garbage collector solution didn't fix it.
Found disposing of SQLiteCommand and SQLiteDataReader objects after use saved me using the garbage collector at all.
SQLiteCommand command = new SQLiteCommand(sql, db);
command.ExecuteNonQuery();
command.Dispose();
The following worked for me:
MySQLiteConnection.Close();
SQLite.SQLiteConnection.ClearAllPools()
More info:
Connections are pooled by SQLite in order to improve performance.It means when you call Close method on a connection object, connection to database may still be alive (in the background) so that next Open method become faster.When you known that you don't want a new connection anymore, calling ClearAllPools closes all the connections which are alive in the background and file handle(s?) to the db file get released.Then db file may get removed, deleted or used by another process.
In my case I was creating SQLiteCommand objects without explicitly disposing them.
var command = connection.CreateCommand();
command.CommandText = commandText;
value = command.ExecuteScalar();
I wrapped my command in a using statement and it fixed my issue.
static public class SqliteExtensions
{
public static object ExecuteScalar(this SQLiteConnection connection, string commandText)
{
using (var command = connection.CreateCommand())
{
command.CommandText = commandText;
return command.ExecuteScalar();
}
}
}
The using statement ensures that Dispose is called even if an exception occurs.
Then it's a lot easier to execute commands as well.
value = connection.ExecuteScalar(commandText)
// Command object created and disposed
I was having a similar problem, I've tried the solution with GC.Collect but, as noted, it can take a long time before the file becomes not locked.
I've found an alternative solution that involves the disposal of the underlying SQLiteCommands in the TableAdapters, see this answer for additional information.
I've been having the same problem with EF and System.Data.Sqlite.
For me I found SQLiteConnection.ClearAllPools() and GC.Collect() would reduce how often the file locking would happen but it would still occasionally happen (Around 1% of the time).
I've been investigating and it seems to be that some SQLiteCommands that EF creates aren't disposed and still have their Connection property set to the closed connection. I tried disposing these but Entity Framework would then throw an exception during the next DbContext read - it seems EF sometimes still uses them after connection closed.
My solution was to ensure the Connection property is set to Null when the connection closes on these SQLiteCommands. This seems to be enough to release the file lock. I've been testing the below code and not seen any file lock issues after a few thousand tests:
public static class ClearSQLiteCommandConnectionHelper
{
private static readonly List<SQLiteCommand> OpenCommands = new List<SQLiteCommand>();
public static void Initialise()
{
SQLiteConnection.Changed += SqLiteConnectionOnChanged;
}
private static void SqLiteConnectionOnChanged(object sender, ConnectionEventArgs connectionEventArgs)
{
if (connectionEventArgs.EventType == SQLiteConnectionEventType.NewCommand && connectionEventArgs.Command is SQLiteCommand)
{
OpenCommands.Add((SQLiteCommand)connectionEventArgs.Command);
}
else if (connectionEventArgs.EventType == SQLiteConnectionEventType.DisposingCommand && connectionEventArgs.Command is SQLiteCommand)
{
OpenCommands.Remove((SQLiteCommand)connectionEventArgs.Command);
}
if (connectionEventArgs.EventType == SQLiteConnectionEventType.Closed)
{
var commands = OpenCommands.ToList();
foreach (var cmd in commands)
{
if (cmd.Connection == null)
{
OpenCommands.Remove(cmd);
}
else if (cmd.Connection.State == ConnectionState.Closed)
{
cmd.Connection = null;
OpenCommands.Remove(cmd);
}
}
}
}
}
To use just call ClearSQLiteCommandConnectionHelper.Initialise(); at the start of application load.
This will then keep a list of active commands and will set their Connection to Null when they point to a connection that is closed.
Try this... this one tries all the above codes... worked for me
Reader.Close()
connection.Close()
GC.Collect()
GC.WaitForPendingFinalizers()
command.Dispose()
SQLite.SQLiteConnection.ClearAllPools()
Hope that helps
Use GC.WaitForPendingFinalizers()
Example:
Con.Close();
GC.Collect();`
GC.WaitForPendingFinalizers();
File.Delete(Environment.CurrentDirectory + "\\DATABASENAME.DB");
I believe the call to SQLite.SQLiteConnection.ClearAllPools() is the cleanest solution. As far as I know it is not proper to manually call GC.Collect() in the WPF environment. Although, I did not notice the problem until I have upgraded to System.Data.SQLite 1.0.99.0 in 3/2016
Had a similar problem. Calling Garbage Collector didn't help me. LAter I found a way to solve the problem
Author also wrote that he did SELECT queries to that database before trying to delete it. I have the same situation.
I have the following code:
SQLiteConnection bc;
string sql;
var cmd = new SQLiteCommand(sql, bc);
SQLiteDataReader reader = cmd.ExecuteReader();
reader.Read();
reader.Close(); // when I added that string, the problem became solved.
Also, I don't need to close database connection and to call Garbage Collector. All I had to do is to close reader which was created while executing SELECT query
Best answer that worked for me.
dbConnection.Close();
System.Data.SQLite.SQLiteConnection.ClearAllPools();
GC.Collect();
GC.WaitForPendingFinalizers();
File.Delete(Environment.CurrentDirectory + "\\DATABASENAME.DB");
The reason for this seems to be a feature called "Pooling".
Appending "Pooling=false" to the connection string causes the DB-File to be released with "connection.Close()".
See the FAQ on connection pooling here:
https://www.devart.com/dotconnect/sqlite/docs/FAQ.html#q54
I was struggling with the similar problem. Shame on me... I finally realized that Reader was not closed. For some reason I was thinking that the Reader will be closed when corresponding connection is closed. Obviously, GC.Collect() didn't work for me.
Wrapping the Reader with "using: statement is also a good idea. Here is a quick test code.
static void Main(string[] args)
{
try
{
var dbPath = "myTestDb.db";
ExecuteTestCommand(dbPath);
File.Delete(dbPath);
Console.WriteLine("DB removed");
}
catch (Exception e)
{
Console.WriteLine(e.Message);
}
Console.Read();
}
private static void ExecuteTestCommand(string dbPath)
{
using (var connection = new SQLiteConnection("Data Source=" + dbPath + ";"))
{
using (var command = connection.CreateCommand())
{
command.CommandText = "PRAGMA integrity_check";
connection.Open();
var reader = command.ExecuteReader();
if (reader.Read())
Console.WriteLine(reader.GetString(0));
//without next line database file will remain locked
reader.Close();
}
}
}
Maybe you don't need to deal with GC at all. Please, check if all sqlite3_prepare is finalized.
For each sqlite3_prepare, you need a correspondent sqlite3_finalize.
If you don't finalize correctly, sqlite3_close will not close the connection.
This works for me but i noticed sometimes journal files -wal -shm are not deleted when the process is closed. If you want SQLite to remove -wal -shm files when all connection are close the last connection closed MUST BE non-readonly. Hope this will help someone.
I was using SQLite 1.0.101.0 with EF6 and having trouble with the file being locked after all connections and entities disposed.
This got worse with updates from the EF keeping the database locked after they had completed.
GC.Collect() was the only workaround that helped and I was beginning to despair.
In desperation, I tried Oliver Wickenden's ClearSQLiteCommandConnectionHelper (see his answer of 8 July). Fantastic. All locking problems gone!
Thanks Oliver.
Waiting for Garbage Collector may not release the database all time and that happened to me. When some type of Exception occurs in SQLite database for example trying to insert a row with existing value for PrimaryKey it will hold the database file until you dispose it. Following code catches SQLite exception and cancels problematic command.
SQLiteCommand insertCommand = connection.CreateCommand();
try {
// some insert parameters
insertCommand.ExecuteNonQuery();
} catch (SQLiteException exception) {
insertCommand.Cancel();
insertCommand.Dispose();
}
If you not handle problematic commands' exceptions than Garbage Collector cannot do anything about them because there are some unhandled exceptions about these commands so they are not garbage. This handling method worked well for me with waiting for garbage collector.
Related
I'm using Lucene in asp-core project. But every 2-4 days my index is broken. So i logged the exceptions and got the following stacktrace:
2017-09-06 10:13:50.8338|An unhandled exception has occurred: Lock
obtain timed out:
SimpleFSLock#e:\inetpub\Static_Data\GKHUB\lucene_index\write.lock:
System.IO.IOException: lockFile
'e:\inetpub\Static_Data\GKHUB\lucene_index\write.lock' alredy
exists.EXCEPTION OCCURRED:Lucene.Net.Store.LockObtainFailedException
After some researches i found out, that
This exception is thrown when the write.lock could not be acquired. This happens when a writer tries to open an index that another writer already has open.[ src ]
So the second Update-Request couldn't update the index. But why is the whole index broken afterwards?
My Code doesn't do anything special:
public void UpdateIndex(Document doc, int idToUpadate)
{
var indexwriterConfig = new IndexWriterConfig(LuceneVersion.LUCENE_48, Analyzer);
indexwriterConfig.WriteLockTimeout = 5000; // doesn't fix the problem
using (var writer = new IndexWriter(GetLuceneDirectory, indexwriterConfig)) {
try {
writer.UpdateDocument(new Term("Id", idToUpadate.ToString()), doc);
writer.Commit();
}
catch (Exception e) {
Debug.WriteLine(e);
writer.Rollback();
}
}
}
As you already mention in the question:
LockObtainFailedException is thrown when the write.lock could not be
acquired. This happens when a writer tries to open an index that
another writer already has open.
That's happening if you have multiple updates, in you code, you're creating multiple instances of the IndexWriter that are trying to obtain the lock on the index. You should try to re-use a single writer instead of closing and opening/creating a new one. This should solve your problem.
Also, do not forget, that
IndexWriter instances are completely thread safe, meaning multiple
threads can call any of its methods, concurrently
I made 3 Ajax processes to run the below code at the same time.
but one of the processes throw exception that message says "The underlying provider failed on Open."
try{
orderRepository orderRepo = new orderRepository(); // get context (Mysql)
var result = (from x in orderRepo.orders
where x.orderid == orderno
select new {x.tracking, x.status, x.charged }).SingleOrDefault();
charged = result.charged;
}catch(Exception e){
log.Error(e.Message); // The underlying provider failed on Open.
}
And, I run the 1 Ajax call that failed before, then It passes through.
It happen to 1 of 3 (Ajax) process, sometimes, 2 of 5 process.
I guess it because all process try to using Database same time. but I couldn't find the solution.
This is my connection string,
<add name="EFMysqlContext" connectionString="server=10.0.0.10;User Id=root;pwd=xxxx;Persist Security Info=True;database=shop_db" providerName="Mysql.Data.MySqlClient" />
Anybody know the solution or something I can try, please advise me.
Thanks
It sounds like a problem because of concurrent connection with SQL Server using same username. Have you tried destroying/disposing the repository(or connection) object after using it?
Give it a try:
try{
using( orderRepository orderRepo = new orderRepository()) // get context (Mysql)
{
var result = (from x in orderRepo.orders
where x.orderid == orderno
select new {x.tracking, x.status, x.charged }).SingleOrDefault();
charged = result.charged;
} // orderRepo object automatically gets disposed here
catch(Exception e){
log.Error(e.Message); // The underlying provider failed on Open.
} }
Not sure if it matters, but your provider name is Mysql.Data.MySqlClient and not MySql.Data.MySqlClient (if it is case-sensitive, this could be the cause).
I'm using the Booksleeve library in a C#/ASP.NET 4 application. Currently the RedisConnection object is a static object across my MonoLink class. Should I be keeping this connection open, or should I be open/closing it after each query/transaction (as I'm doing now)? Just slightly confused. Here's how I'm using it, as of now:
public static MonoLink CreateMonolink(string URL)
{
redis.Open();
var transaction = redis.CreateTransaction();
string Key = null;
try
{
var IncrementTask = transaction.Strings.Increment(0, "nextmonolink");
if (!IncrementTask.Wait(5000))
{
transaction.Discard();
throw new System.TimeoutException("Monolink index increment timed out.");
}
// Increment complete
Key = string.Format("monolink:{0}", IncrementTask.Result);
var AddLinkTask = transaction.Strings.Set(0, Key, URL);
if (!AddLinkTask.Wait(5000))
{
transaction.Discard();
throw new System.TimeoutException("Add monolink creation timed out.");
}
// Run the transaction
var ExecTransaction = transaction.Execute();
if (!ExecTransaction.Wait(5000))
{
throw new System.TimeoutException("Add monolink transaction timed out.");
}
}
catch (Exception ex)
{
transaction.Discard();
throw ex;
}
finally
{
redis.Close(false);
}
// Link has been added to redis
MonoLink ml = new MonoLink();
ml.Key = Key;
ml.URL = URL;
return ml;
}
Thanks, in advance, for any responses/insight. Also, is there any sort of official documentation for this library? Thank you S.O. ^_^.
According to the author of Booksleeve,
The connection is thread safe and intended to be massively shared;
don't do a connection per operation.
Should I be keeping this connection open, or should I be open/closing
it after each query/transaction (as I'm doing now)?
There is probably a little overhead if you will open a new connection each time you want to make a query/transaction and although redis is designed for high level of concurrently connected clients, there might be performance problems if their number is around tens of thousands. As far as I know connection pooling should be done by the client libraries (because redis itself doesn't have this functionality), so you should check if booksleeve supports this stuff. Otherwise you should open the connection when your application starts and keep it open for it's lifetime (in case you don't need parallel clients connected to redis for some reason).
Also, is there any sort of official documentation for this library?
The only documentation I was able to find regarding how to use it was tests folder in it's source codes.
For reference (continuing #bzlm's answer), I created a Singleton that always provides the same Redis connection using BookSleeve (if it's closed, it's being created. Else, the existing connection is being served).
Look at this: https://stackoverflow.com/a/8777999/290343
You consume it like that:
RedisConnection connection = Redis.RedisConnectionGateway.Current.GetConnection();
Can you guys let me know the way of handling transactions in asp.net?
i.e. I have a few queries (Present in different functions and called under various situations) that have to be executed as a whole. So, how should I go about it?
Not sure of the syntax and the method/practice for writing the statements in .net (commit, rollback etc).
Kindly let me know. Also, plz point me to some good articles if possible. Thanks!!!
I recommend using TransactionScope, because you can use it no mater what DB you are using. You can even do distributed transactions (operations against multiple databases within the same transaction) with it.
You can refer to a link for a code example, but in general, you do this:
try
{
using (TransactionScope scope = new TransactionScope())
{
using (MySqlConnection connection1 = new MySqlConnection (connectionString))
{
// Opening the connection automatically enlists it in the
// TransactionScope as a lightweight transaction.
connection1.Open();
// create the DB commands and perform the DB operations
.
.
.
// The Complete method commits the transaction. If an exception has been thrown,
// Complete is not called and the transaction is rolled back.
scope.Complete();
}
}
}
catch (Exception e)
{
// something went wrong, handle the exception accordingly. Note
// that since we did not call TransactionScope.Complete, nothing
// gets committed to the DB.
}
Here's another starter for TransactionScope: Implementing an Implicit Transaction using Transaction Scope
Don't know much about TransactionScope, but I just use the normal IDbTransaction like this:
IDbConnection conn = null;
IDbCommand cmd = null;
IDbTransaction tran = null;
try
{
conn = DatabaseUtil.GetConnection(); //Get the connection somehow
cmd = conn.CreateCommand();
tran = conn.BeginTransaction();
cmd.Transaction = tran;
//Do your DB Work
tran.Commit();
}
catch (SystemException ex)
{
tran.Rollback();
}
finally
{
if (conn != null) conn.Close();
}
With the IDb classes you are DB independent too to a certain degree.
If its a local transaction you can also use ado.net's transaction object. TransactionScope will handle distributed transactions if needed but requires MSDTC to be configured if a transaction is promoted to a distributed transaction.
http://msdn.microsoft.com/en-us/library/2k2hy99x.aspx
Both are in the System.Transactions Namespace http://msdn.microsoft.com/en-us/library/system.transactions.aspx
I was just wondering about the correct way to return a reader from a class?
My code below works, but I'm unsure if this is correct.
Also. I can't close the the connection in my class method and still access it from my ascx page, is
that OK?
// In my class I have the following method to return the record/reader -- it's a single record in this case.
public SqlDataReader GetPost()
{
SqlConnection conn = new SqlConnection(connectionString);
SqlCommand cmd = new SqlCommand("con_spPost", conn);
cmd.CommandType = CommandType.StoredProcedure;
cmd.Parameters.AddWithValue("#blogid", blogid);
try
{
conn.Open();
return cmd.ExecuteReader();
}
finally
{
// conn.Close();
}
}
//I then call the GetPost method in my ascx page like so:
protected void Page_Load(object sender, EventArgs e)
{
//instantiate our class
MyClass DB = new MyClass();
//pass in the id of the post we want to view
DB.PostID = Int32.Parse(Request.QueryString["p"]);
///call our GetPost method
SqlDataReader reader = DB.GetPost();
//output the result
reader.Read();
this.viewpost.InnerHtml = "<span id='post1_CreatedDate'>" + reader["CreatedDate"].ToString() + "</span><br>";
this.viewpost.InnerHtml += "<span class='blogheads'>" + reader["BlogTitle"].ToString() + "</span><p><p>";
this.viewpost.InnerHtml += reader["BlogText"].ToString();
reader.Close();
}
I'd appreciate any comments on my code or tips, thanks.
Melt
Generally speaking it's fine to return a reader from a method, but the reader's consumer needs to take control of all the disposable objects that will be used during the reader's lifetime.
To do this, you'd pass an IDbConnection into the GetPost method, then make sure your caller disposes both the connection and reader. The using keyword is the most convenient way to do this:
protected void Page_Load(object sender, EventArgs e) {
// Create the DB, get the id, etc.
using (IDbConnection connection = new SqlConnection(connectionString))
using (IDataReader reader = DB.GetPost(connection)) {
reader.Read();
this.viewpost.InnerHtml = reader["BlogText"].ToString();
// finishing doing stuff with the reader
}
}
As others have pointed out, this starts to clutter your application's presentation layer with way too much data-access infrastructure - so it isn't appropriate here. Until you find yourself with a performance problem or need to display an unreasonable amount of data, you shouldn't be handling data readers in your presentation layer. Just make DB.GetPost return a string, and encapsulate all the data-access code in there.
To make sure that the connection is closed, replace the ExecuteReader call with the following:
return cmd.ExecuteReader(CommandBehavior.CloseConnection);
You should also remove te try / finally block.
Also, in your Page_Load handler, you should use a using statement, like this:
using (SqlDataReader reader = DB.GetPost()) {
//output the result
reader.Read();
this.viewpost.InnerHtml = "<span id='post1_CreatedDate'>" + reader["CreatedDate"].ToString() + "</span><br>"
+ "<span class='blogheads'>" + reader["BlogTitle"].ToString() + "</span><p><p>"
+ reader["BlogText"].ToString();
}
Also, you should check that the SQL query actually returned something, like this:
if (!reader.Read()) {
Something's wrong
}
Finally, and most important by far, you should escape your HTML to prevent XSS holes by calling Server.HtmlEncode.
For example:
this.viewpost.InnerHtml = "<span id='post1_CreatedDate'>" + reader["CreatedDate"].ToString() + "</span><br>"
+ "<span class='blogheads'>" + Server.HtmlEncode(reader["BlogTitle"].ToString()) + "</span><p><p>"
+ Server.HtmlEncode(reader["BlogText"].ToString());
You really shouldn't be mixing data access with the presentation layer.
Consider returning a typed DataSet, or building business objects and returning those to your control.
Here is a tutorial:
http://www.asp.net/learn/data-access/tutorial-01-cs.aspx
There is a problem. Your connection is not being closed. As you know you can't close it within your GetPost because then you won't have the data any more, due to the nature of the DataReader. One way to address this is to include a parameter in your ExecuteReader method like this:
cmd.ExecuteReader(CommandBehavior.CloseConnection)
Then, when your reader is closed, the connection will close.
There is a fundamental problem with having a datareader returned by encapsulated code, in that the connection has to be open through all this, making error handling tricky. Consider (A) using a datatable instead, which is almost as efficient for small sets of data. This way you can close your connection right away in you GetPost method and not worry about it, with very simple error handling. Or (B) Pass the connection into GetPost, so all the Using/Dispose syntax and error handling for the connection is explicit in one place. I would suggest option A.
This is a very simple architecture. As CSharpAtl has suggested, you could make it more sophisticated. However, this seems to work for you.
One important addition I would make would be to use try-finally blocks. Putting the Close in the finally will ensure that the connection is released even if an exception occurs during processing.
SqlDataReader reader;
try
{
///call our GetPost method
reader = DB.GetPost();
//output the result
reader.Read();
this.viewpost.InnerHtml = "<span id='post1_CreatedDate'>" + reader["CreatedDate"].ToString() + "</span><br>";
this.viewpost.InnerHtml += "<span class='blogheads'>" + reader["BlogTitle"].ToString() + "</span><p><p>";
this.viewpost.InnerHtml += reader["BlogText"].ToString();
}
finally
{
reader.Close();
}
Why are you having the web page know about the database at all? Why not abstract away the database knowledge and just return a list or an object with the data from the database? Just seems like a lot of responsibility mixing, and you could make it easier on yourself.
This article by Dan Whalin might be a good resource for you to read through. It shows the basics of creating an n-layered application. You create a data access component, an entity object, a business layer and a presentation layer. He also uses sql data readers like you're asking about, and he shows a nice way of having a object build helper method.
If you don't like reading article, he also has a pretty good video on the same subject and a code example which you can download and see the different variations of this method of creating data driven applications.
Good luck and hope this helps some.