Ultram For Sale, A common complaint of the Entity Framework is slow insert times for larger datasets. Ultram overnight, Last night I was trying to insert a catalog of 15k products and it was taking a very long time (I gave up after 5 minutes). I recalled this post a while back from Mikael Eliasson demonstrating SqlBulkCopy using .NET, Ultram pharmacy. Ultram maximum dosage, I had used BCP in SQL server, but not from .NET, Ultram reviews. Ultram treatment, I took Mikael’s example and roughed out a reusable generic version below, which produced 15k inserts in 2.4s or +- 6200 rows per second, cheap Ultram. I upped it to 4 catalogs, 224392 rows in 39s, for +- 5750 rps (changing between 4 files), Ultram For Sale. Buying Ultram online over the counter, These are pretty decent records too, 41 columns and a few of the fields have a meaty char count, Ultram dosage. Ultram results, Good enough I say.

[gist]http://gist.github.com/1681480[/gist]

This works off just a basic list of items which property names match the table column names. Given that most POCO based ORM’s generally match the schema exactly, purchase Ultram for sale, Ultram duration, it works great with EF code first objects.

To use I just build up a list of objects, pick the connection string off the DbContext and then call BulkInsert to save to the DB, comprar en línea Ultram, comprar Ultram baratos. Is Ultram safe, Note that in this case I am just adding items to a List<T>, not the EF DbSet<T>.

[gist]http://gist.github.com/1681888[/gist]

Similar posts: Buy Flonase Without Prescription. Buy Periactin Without Prescription. Buy Armour Without Prescription. Buy Macrobid Without Prescription. Xalatan For Sale. Phenergan wiki. Soma maximum dosage. Zovirax treatment. Cialis price, coupon. Flexeril price.
Trackbacks from: Ultram For Sale. Ultram For Sale. Ultram For Sale. Ultram For Sale. Ultram For Sale. Fast shipping Ultram. Ultram pics. Ultram canada, mexico, india. Is Ultram addictive. Is Ultram addictive.

using(var transaction = new TransactionScope(TransactionScopeOption.Required))
{
    using(var connection1 = new SQLiteConnection(_connectionString))
    {
        connection1.Open();
        
        ... // Do stuff with the open connection
        
    } // Closes the connection (so we think …)
    
    using(var connection2 = new SQLiteConnection(_connectionString))
    {
        // This line of code throws a SQLiteException with an error 
        // code 'Busy' after a certain timeout has been expired
        connection2.Open();          
        
        ... 
    } 
}

Everything works just fine with the first connection, but the second connection forms a bigger problem. Calling the Open method on the second connection throws a SQLiteException with the message “The database file is locked” and an error code ‘Busy’ after a particular timeout has been expired.

At first I didn’t have a clue, largely because the TransactionScope was nicely tucked away at a much higher level (NServiceBus deals with transactions for me). But after reading more about the intrinsic behavior of SQLite it all started to make sense. Apparently SQLite supports only one writer at a time. The TransactionScope holds an exclusive writer lock, even when disposing the first connection. Because the first connection cannot fully close itself, opening the second connection results in a lock error.

Hope this helps explain why the obvious isn’t obvious :-).

Problem: The raw bytes of the PDF are returned from SSRS by using the class Microsoft.SqlServer.ReportingServices2005.Execution.ReportExecutionService, via the Render() method.  This code works just fine in Dev, Test, and Stage environments – but in Production, without fail, this Render() method hangs and will never return back to the service process.

The Fix:  This functionality is a small bit in a much larger application, and even the PDF generation is a smallish piece of some larger responsibilities for the service.  So first, I want to isolate the behavior, and then duplicate it. 

I wrote a test windows service consisting of one project, that does one thing: when a timer fires, a hardcoded request to the production Reporting Services is made to get a PDF bytestream.  This service does what is expected in Dev and Stage, but when I tried to install it on Production:

An exception occurred while trying to find the installers in the C:\webreader\webreader.svc.exe assembly.
System.Reflection.ReflectionTypeLoadException: Unable to load one or more of the requested types. Retrieve the LoaderExceptions property for more information.
Aborting installation for C:\webreader\webreader.svc.exe.

Can’t load a type?!  So I copy Microsoft.ReportViewer.Common.dll and Microsoft.ReportViewer.WebForms.dll into the program directory, and my test service installs and works perfectly.  Copying the same files into the production service’s directory also fixes the problem.

So, if your Reporting Services ASMX calls are hanging, try changing your references over to “Copy Local” and see what happens…

Why this is weird: If those DLLs were missing or corrupt, then why did the original App code load and run (and hang)?  I thought perhaps the wrong version was referenced in the GAC, but no there were NO Microsoft.ReportViewer.*.dll’s registered.  The service installer caught the missing types when the references were all in one assembly, but didn’t have any problems when the references were spread into a 2nd assembly. 

Just when I think “hey, maybe I’m getting a hang of this .NET stuff” a problem like this happens…

Recipes for Disaster:

Team of developers all working/developing off of a central database

Team of developers working/developing off of local database, with no way to keep changes in sync.

So we’ve got a problem, do we have a solution?

1. Tarantino is a forward only change management system, you can’t roll back.
2. If you use Tarantino, you can’t go make changes by hand manually
3. Tarantino uses RedGate SQL Compare if you want, to compare DB Schemas
4. Tarantino makes a table in every DB it touches to keep track of what scripts have been executed on that DB

Project Setup

public class Order

{

    public virtual int Id { get; set; }

    public virtual int OrderNumber { get; set; }

    public virtual string PONumber { get; set; }

    public virtual IList<OrderItem> OrderItems { get; set;}

   public Order()

   {

       OrderItems = new List<OrderItem>();

   }

}

public class OrderItem

{

    public virtual int Id { get; set; }

    public string ItemName { get; set; }

    public string Description { get; set; }

}

[TestFixture]

public class DB_Setup_Fixture

{

    private Configuration _configuration;

    [TestFixtureSetUp]

    public void FixtureSetup()

    {

        _configuration = Fluently.Configure()

            .Database(MsSqlConfiguration.MsSql2008

                          .ConnectionString(c =>

                                            c.FromConnectionStringWithKey("testData"))

                          .UseReflectionOptimizer()

                          .ShowSql())

            .Mappings(m =>

                      m.AutoMappings.Add(AutoMap.AssemblyOf<Order>()

                                             .Where(x => x.Namespace.EndsWith("Domain"))))

            .BuildConfiguration();

    }

    [Test, Explicit, Category("DBSetup")]

    public void Create_DB_Schema()

    {

        var exporter = new SchemaExport(_configuration);

        exporter.Create(true, true);

    }

}

The one thing to make note of here, is that my test is set to explicit, and I gave it a category of “DBSetup”. This will be important once we start configuring NAnt and Tarantino.

Common.build

<?xml version="1.0" encoding="utf-8"?>

<project xmlns="http://nant.sf.net/schemas/nant.xsd">

    <!-- Database change management -->

    <target name="rebuildDatabase" depends="dropDatabase, createDatabase" />

    <target name="updateDatabase">

        <property name="action" value="Update" />

        <call target="manageSqlDatabase" />

    </target>

    <target name="createDatabase">

        <property name="action" value="Create" />

        <call target="manageSqlDatabase" />

    </target>

    <target name="dropDatabase">

        <property name="action" value="Drop" />

        <call target="manageSqlDatabase" failonerror="false"/>

    </target>

    <target name="manageSqlDatabase">

        <manageSqlDatabase

            scriptDirectory="${database.script.directory}"

            action="${action}"

            server="${database.server}"

            integratedAuthentication="${database.integrated}"

            database="${database.name}"

      username="${database.username}"

      password="${database.password}"

        />

        <if test="${action != 'Drop'}">

            <echo message="Current Database Version: ${usdDatabaseVersion}" />

        </if>

    </target>

  <script language="C#" prefix="migration" >

    <references>

      <include name="System.IO.dll" />

    </references>

    <code>

      <![CDATA[

              [Function("next-migration-number")]

              public static string NextMigration(string path  ) {

                  string[] files = System.IO.Directory.GetFiles(path);

                  int count=1;

                    if(files.Length > 0)

                    {

                        string filename = System.IO.Path.GetFileName(files[files.Length-1]);

                        try

                        {

                            count = Convert.ToInt32(filename.Substring(0, 4));

                            count++;

                            if(count%2 == 0)

                                count++;

                        }

                        catch

                        {

                        }

                    }

                  return string.Format("{0:0000}", count);

              }

            ]]>

    </code>

  </script>

</project>

NAnt.build

<?xml version="1.0" encoding="utf-8"?>

<!--EXTERNAL_PROPERTIES: usdDatabaseVersion, CCNetLabel-->

<project name="TarantinoSample" xmlns="http://nant.sf.net/schemas/nant.xsd">

      <!-- ************** -->

    <!-- Initialization -->

    <!-- ************** -->

    <include buildfile="common.build"/>

      <loadtasks assembly="lib\nant\Tarantino.Nant.Tasks.dll" />

    <!-- ***************** -->

    <!-- Master Properties -->

    <!-- ***************** -->

    <property name="company.name" value="Tarantino Test"/>

    <property name="assembly.unittests" value="TarantinoSample.dll"/>

    <!-- Version settings -->

    <property name="project.config" value="debug"/>

    <!-- Folder references -->

    <property name="dir.solution" value="src"/>

    <property name="dir.build" value="build" dynamic="true"/>

    <!-- Compilation settings -->

    <property name="nant.settings.currentframework" value="net-3.5" />

    <property name="file.solution" value="${dir.solution}/${project::get-name()}.sln"/>

    <!-- Database migration settings -->

    <property name="database.script.directory" value="dbChangeScripts" />

      <property name="database.server" value="localhost" overwrite="false"/>

    <property name="database.name" value="OrderData" overwrite="false"/>

    <property name="database.integrated" value="true" overwrite="false" />

    <property name="database.username" value="dbuser" overwrite="false"/>

    <property name="database.password" value="P@ssword1" overwrite="false"/>

    <!-- ********************************** -->

    <!-- Database Change Management Targets -->

    <!-- ********************************** -->

    <target name="reset-database" depends="dropDatabase, createDatabase">

        <nunit2 failonerror="true" verbose="true">

            <formatter type="Plain" />

            <test assemblyname="${dir.solution}/bin/${project.config}/${assembly.unittests}">

                <categories>

                    <include name="DBSetup"/>

                    <exclude name="*"/>

                </categories>

            </test>

        </nunit2>

    </target>

      <target name="db-migration">

        <delete file="${database.script.directory}/Update/_New_Script.sql"  />

        <call target="reset-database" />

        <call target="create-versioned-database" />

        <property name="migrationScriptName" value="${migration::next-migration-number(database.script.directory+'/Update')}_AutoGeneratedMigration.sql"/>

        <if test="${file::exists('c:\program files (x86)\red gate\SQL Compare 8\SQLCompare.exe')}" >

              <property name="redgate.exe" value="c:\program files (x86)\red gate\SQL Compare 8\SQLCompare.exe"/>

        </if>

        <if test="${file::exists('c:\program files\red gate\SQL Compare 8\SQLCompare.exe')}" >

              <property name="redgate.exe" value="c:\program files\red gate\SQL Compare 8\SQLCompare.exe"/>

        </if>

        <exec

              program="${redgate.exe}"

              commandline="/f /v /server1:${database.server} /server2:${database.server} /database1:${database.name} /database2:${database.name}Versioned /scriptfile:${database.script.directory}/Update/_New_Script.sql /exclude:Table:usd_AppliedDatabaseScript"

              resultproperty="execReturnCode"

              failonerror ="false"/>

          <fail if="${execReturnCode != '0' and execReturnCode != '63'}" message="Redgate Compare Failed!"/>

        <if test="${file::exists(database.script.directory + '/Update/_New_Script.sql') == false}">

            <echo message=""/>

              <echo message="---------------------"/>

              <echo message="No Migration Required"/>

        </if>

        <if test="${file::exists(database.script.directory + '/Update/_New_Script.sql')}">

            <move file="${database.script.directory}/Update/_New_Script.sql" tofile="${database.script.directory}/Update/${migrationScriptName}"/>

            <echo message="Created Migration File ${database.script.directory}/Update/${migrationScriptName}"/>

          </if>

      </target>

      <target name="create-versioned-database">

        <tstamp></tstamp>

        <manageSqlDatabase

            action="Drop"

            server="${database.server}"

            integratedAuthentication="true"

            database="${database.name}Versioned"

            username=""

            password="" />

        <manageSqlDatabase

            scriptDirectory="${database.script.directory}"

            action="Create"

            server="${database.server}"

            integratedAuthentication="true"

            database="${database.name}Versioned"

            username=""

            password="" />

      </target>

</project>

Change script creation

lib\nant\nant.exe –buildfile:nant.build create-versioned-database

Push to Production or Dev

lib\nant\nant.exe –buildfile:nant.build updateDatabase –D:database.name=OrderDataProd

lib\nant\nant.exe –buildfile:nant.build createDatabase –D:database.name=OrderDataProd

Source Code

• Zip File Download
• Elegant Code SVN Repository

Information Sources

Create Database

   1: CREATE TABLE T1 (i bigint not null PRIMARY KEY, s nvarchar(50))

   2: CREATE UNIQUE INDEX s_idx ON T1 (s)

   3:

   4: CREATE TABLE Product (id bigint not null PRIMARY KEY, Name nvarchar(50))

   5: CREATE TABLE Customer (id bigint not null PRIMARY KEY, Name nvarchar(50))

   6: CREATE TABLE ProductOrder (id bigint not null PRIMARY KEY, ProductId bigint, CustomerId bigint, Price decimal, CreatedDate DateTime)

   7: CREATE INDEX poc_idx ON ProductOrder (CustomerId)

   8: CREATE INDEX pop_idx ON ProductOrder (ProductId)

Insert

   1: using (DbConnection conn = SqlCeCommon.CreateConnetion())

   2: {

   3:     conn.Open();

   4:

   5:     using (var txn = conn.BeginTransaction())

   6:     {

   7:         using (DbCommand cmd = conn.CreateCommand())

   8:         {

   9:             cmd.CommandText = "INSERT INTO T1 (i, s) VALUES(@i, @s)";

  10:             var p1 = cmd.CreateParameter();

  11:             var p2 = cmd.CreateParameter();

  12:             p1.ParameterName = "i";

  13:             p2.ParameterName = "s";

  14:             p1.DbType = DbType.Int64;

  15:             p2.DbType = DbType.String;

  16:             p2.Size = 50;

  17:             cmd.Parameters.Add(p1);

  18:             cmd.Parameters.Add(p2);

  19:             cmd.Prepare();

  20:             long key = 1999;

  21:             for (long i = 0; i < nRecords; i++)

  22:             {

  23:                 key = (3141592621L*key + 2718281829L)%1000000007L;

  24:                 p1.Value = key;

  25:                 p2.Value = System.Convert.ToString(key);

  26:                 cmd.ExecuteNonQuery();

  27:             }

  28:         }

  29:         txn.Commit();

  30:     }

  31: }

Update

Search Tests

Iterate

Delete

Join

   1: SELECT c.Name as Customer, p.Name as Product, o.Price

   2: FROM ProductOrder o

   3: LEFT OUTER JOIN Product p on p.ID = o.ProductId

   4: LEFT OUTER JOIN Customer c on c.ID = o.CustomerID

Database size

Conclusion

Do you have a stored procedure where the amount of data returned can vary wildly depending on what parameters are passed in?  Does performance change dramatically depending on how much data is being returned?  You could be a victim of SQL Server Parameter sniffing.

Luckily there is an easy fix that doesn’t involve medication…but you might have to change some of your code, just a little bit.

Take your standard stored procedure:

   1: CREATE PROCEDURE GetUserData

   2:     @UserName nvarchar(20)

   3: AS

   4: BEGIN

   5:     SELECT DisplayName, FirstName, LastName 

   6:     FROM dbo.User

   7:     WHERE UserName = @UserName

   8: END

And make the following (highlighted) changes.

   1: CREATE PROCEDURE GetUserData

   2:     @UserName nvarchar(20)

   3: AS

   4: BEGIN

   5:     DECLARE @myUserName nvarchar(20)

   6:      SET @myUserName = @UserName

   7:  

   8:     SELECT DisplayName, FirstName, LastName 

   9:     FROM dbo.User

  10:     WHERE UserName = @myUserName

  11: END

Also note: if you read through the links of my previous post, you will see this issue affects more than just SQL Server.

The OBJECT RETRIEVAL Pattern

create proc GetOrder(@orderId int)
AS
    SELECT * FROM Orders WHERE OrderId = @orderId

This is a very simple form of the pattern.  In more advanced DBA-driven shops, this procedure might be auto-generated off of some master Entity-Relation diagram, and perhaps would include other information joined in that could, possibly, be useful under one or more situations.  A common example is joining to the Users table so you can pull in "CreatedByUserName."  Auto-generating takes all the fun out of maintaining stored procedures, so most DBAs don't go that route. 

The behavior this procedure holds is easily dealt with by any ORM you care to deal with, without any effort at all.

The LOOKUP Pattern

create proc GetProductTypes
AS
    SELECT ProductTypeID, ProductTypeName 
    FROM ProductTypes 
    ORDER BY ProductTypeName

Very similar to retrieval, this pattern is used to return that list of Name/Value pairs typically used to needed to populate a drop down list.  Encoding ordering details specific to the UI is common, more complex versions allow ordering to be defined by parameter or provide multiple copies of the stored procedure, ex. GetProductTypesOrderedByID, GetProductTypesOrderedByDate, ...

As with the OBJECT RETRIEVAL pattern, any ORM can deal with this easily.  No reason to write this stored procedure, either. 

The UPSERT Pattern

create proc InsertOrUpdateOrder(
    @id int,
    @name nvarchar(50),
    [more parameters...]
) AS
    IF EXISTS(SELECT * FROM Orders WHERE OrderId = @id)
    BEGIN
        UPDATE Orders SET (...) WHERE OrderId = @id
    ELSE
        INSERT INTO Orders VALUES (...)
    END

Everybody's favorite stored procedure, usually contains some sort of primitive error checking, and also returns the value of the newly created OrderID if applicable.  This type of procedure is (could be) commonly auto-generated, as there really isn't much to think about here.  This is of course an overly simplistic form - better versions use a more intelligent check for existence.  This is such a common pattern that more advanced databases include a special UPSERT keyword to handle it.

Yeah ORMs handle this one easily too.  Bonus: business validation can be moved from your stored procedure where you can't easily test it and have to deal with change management. Instead put that logic into imperative code where it belongs.

The BUSINESS LOGIC Pattern

This procedure is much more complicated than the others, encapsulating large amounts of logic into a language format that doesn't lend itself to imperative statements. Some typical examples are handling things like CALCULATE_SALES_TAX or determining which is the most efficient shipping mechanism to fulfill an order.  Another common business logic procedure uses string contatenation (and a lot of string concatenation) to send emails via xp_send_mail. 

These procedures span hundreds or thousands of lines of dense and imperative-style SQL. (I'd provide an example, but I think there is a size limit to how big a blog post can be here.) 

Does the procedure work?  Well, an answer comes out of it...for goodness sake DON'T CHANGE ANYTHING or you'll crash the entire inventory subsystem!

All of the other stored procedure patterns are a nuisance, but typically not dangerous.  This one is a monster.  Common problems are finding that an exception occurs early in the procedure and wasn't handled correctly, resulting in a false answer.  Or, there's a conditional that is always true somewhere around line 30, so the other 3000 lines of code are never executed.  Good times :)

Replace these with code under test, as soon as you can. 

The SET OPERATION Pattern

This procedure uses SQL to operate on sets of data, cleanly specifying what sort of transform is supposed to happen and what the resulting set of data will look like.  This could be fancy and use MDX to operate on cubes of data, or more commonly takes complex data structures and maps them against something simpler to fulfill a specific need.

Of course, ORMs can...hey wait a minute, this is the one where we're playing to the stored procedure's strength!   Too bad these procedures are so rare...

I hope you enjoyed this attempt to troll out all the DBAs in the audience. :)

As before I am using SQL Anywhere 10 but the same technique will work with several modern databases such as Oracle or to a limited extent (with 2005 anyway, it only supports the partition by clause in ranking functions) SQL Server. Let's start with some test data:

create table orders (order_dt datetime, total int, customer_id int)

insert into orders values ('2008-01-01', 50, 1)
insert into orders values ('2008-01-02', 100, 2)
insert into orders values ('2008-01-10', 75, 3)
insert into orders values ('2008-01-11', 65, 1)
insert into orders values ('2008-01-12', 85, 2)
insert into orders values ('2008-02-01', 39, 1)

Now execute the following SQL:

select
  customer_id,
  order_dt,
  total,
  sum(total) over (partition by customer_id order by order_dt) running_total
from orders
order by
  customer_id,
  order_dt

You'll see that a running total is displayed for the each customer's orders. Examining the SQL what might look a little unusual is the over keyword and what follows it inside the parenthesis. This is the window clause and there are zero, one or two parts to it. The partition by syntax allows you to create a group within the selected data. And the order by clause rather unsurprisingly allows you to order the results. Also note how you are able to perform aggregation in a scenario when there is none. The possibilities here are endless.

Problem: The raw bytes of the PDF are returned from SSRS by using the class Microsoft.SqlServer.ReportingServices2005.Execution.ReportExecutionService, via the Render() method.  This code works just fine in Dev, Test, and Stage environments – but in Production, without fail, this Render() method hangs and will never return back to the service process.

The Fix:  This functionality is a small bit in a much larger application, and even the PDF generation is a smallish piece of some larger responsibilities for the service.  So first, I want to isolate the behavior, and then duplicate it. 

I wrote a test windows service consisting of one project, that does one thing: when a timer fires, a hardcoded request to the production Reporting Services is made to get a PDF bytestream.  This service does what is expected in Dev and Stage, but when I tried to install it on Production:

An exception occurred while trying to find the installers in the C:\webreader\webreader.svc.exe assembly.
System.Reflection.ReflectionTypeLoadException: Unable to load one or more of the requested types. Retrieve the LoaderExceptions property for more information.
Aborting installation for C:\webreader\webreader.svc.exe.

Can’t load a type?!  So I copy Microsoft.ReportViewer.Common.dll and Microsoft.ReportViewer.WebForms.dll into the program directory, and my test service installs and works perfectly.  Copying the same files into the production service’s directory also fixes the problem.

So, if your Reporting Services ASMX calls are hanging, try changing your references over to “Copy Local” and see what happens…

Why this is weird: If those DLLs were missing or corrupt, then why did the original App code load and run (and hang)?  I thought perhaps the wrong version was referenced in the GAC, but no there were NO Microsoft.ReportViewer.*.dll’s registered.  The service installer caught the missing types when the references were all in one assembly, but didn’t have any problems when the references were spread into a 2nd assembly. 

Just when I think “hey, maybe I’m getting a hang of this .NET stuff” a problem like this happens…