Question

I’m using Entity Framework (4.3) and code first (along with SQL server 2012 localdb).

I’m having an interesting situation where in one application a call returns a result (‘card’) and another using the same query on the same db (with thousands of rows available) returns nothing (null in this case since I’m using FirstOrDefault).

using (var context = new MyEntities())
        {
            //verify account exists
            var account = context.Account.SingleOrDefault(a => a.AccountNum == accountNum);

            if (account == null)
            {   
                //handle no account                 
                return false;
            }                

            var card = context.cards.FirstOrDefault(c => c.Used == false);

            //check that we actually got a card
            if (card == null)
            { 
                //handle no cards available
                return false;
            }

            card.Used = true;

           //snip...

            context.SaveChanges();
        }

In my test console app ‘card’ will have a value, while in my “service” it does not. Same code (same assembly even), same db, same rows available. The account retrieval works in both cases.

I am going to try this:
How to force EF Code First to query the database?

But even if that works the inconsistent behaviour doesn’t make sense to me.

EDIT:

Ends up I have two database instance as pointed out. The interesting part is the datasource changes depending on where the library is run from (as debug under VS11 or from a service running as the admin user). Here’s my app.config:

<entityFramework>
<defaultConnectionFactory type="System.Data.Entity.Infrastructure.SqlConnectionFactory, EntityFramework">
  <parameters>
    <parameter value="Data Source=(localdb)\v11.0; Integrated Security=True; MultipleActiveResultSets=True" />
  </parameters>
</defaultConnectionFactory>

When I attached to the service process running the same library (DLL) as the console application the datasource has changed from:

(localdb)\v11.0

To:

.\SQLEXPRESS

Same DLL, same app.config. I vaguely remember reading something about named vs unamed local db instances and what user they are run under, however my google skills are failing me currently.

So when I open a connection in SSMS I need to use the appropriate server name depending on which db I’m working with. Not a big deal as this is just the setup for development.

My question:

• Does anyone know where this is documented?

Answer 1

The issue was using two databases. Two were being used as I didn’t have a .config in the actual running folder with the connection string for my service. The test application had a .config with

  <entityFramework>
  <defaultConnectionFactory type="System.Data.Entity.Infrastructure.SqlConnectionFactory, EntityFramework">
        <parameters>
             <parameter value="Data Source=(localdb)\v11.0; Integrated Security=True; MultipleActiveResultSets=True" />
        </parameters>
  </defaultConnectionFactory>
  </entityFramework>

The service did not and therefore fell on the default behaviour (which appears is connecting to the default instance of SQL Server).

I would have rather had the application fail to create the DB (throwing an exception) without a connection explicitly stated, however it is documented by Microsoft in the ‘remarks’ section of DbContext:

If no connection string is found, then the name is passed to the DefaultConnectionFactory registered on the Database class. The connection factory then uses the context name as the database name in a default connection string. (This default connection string points to .\SQLEXPRESS on the local machine unless a different DefaultConnectionFactory is registered.)