What does "exec sp_reset_connection" mean in Sql Server Profiler?

Trying to understand what Sql Profiler means by emitting "sp_reset_connection".

I have the following, "exec sp_reset_connection" line followed by BatchStarting and Completed,

RPC:Completed       exec sp_reset_connection
SQL:BatchStarting   SELECT [c].[TestID] AS [TestID], [c].[Description] AS [Description] FROM [dbo].[Test] AS [c]
SQL:BatchCompleted  SELECT [c].[TestID] AS [TestID], [c].[Description] AS [Description] FROM [dbo].[Test] AS [c]    

Basically does first line "exec sp_reset_connection" mean the whole process (my connection was opened, the select stmt is run, then the connection is closed and released back to pool) just take place? Or my connection is still in open stage.

And, why does the sp_reset_connection executed before my own select statement, shouldn't it the reset come after user's sql?

I'm trying to know is there a way to know in more detail when a connection is opened and closed?

By seeing "exec sp_reset_connection", does that mean my connection is closed?

Like the other answers said, sp_reset_connection indicates that connection pool is being reused. Be aware of one particular consequence!

Jimmy Mays' MSDN Blog said:

sp_reset_connection does NOT reset the transaction isolation level to the server default from the previous connection's setting.

UPDATE: Starting with SQL 2014, for client drivers with TDS version 7.3 or higher, the transaction isolation levels will be reset back to the default.

ref: SQL Server: Isolation level leaks across pooled connections

Here is some additional information:

What does sp_reset_connection do?

Data access API's layers like ODBC, OLE-DB and System.Data.SqlClient all call the (internal) stored procedure sp_reset_connection when re-using a connection from a connection pool. It does this to reset the state of the connection before it gets re-used, however nowhere is documented what things get reset. This article tries to document the parts of the connection that get reset.

sp_reset_connection resets the following aspects of a connection:

• All error states and numbers (like @@error)

• Stops all EC's (execution contexts) that are child threads of a parent EC executing a parallel query

• Waits for any outstanding I/O operations that is outstanding

• Frees any held buffers on the server by the connection

• Unlocks any buffer resources that are used by the connection

• Releases all allocated memory owned by the connection

• Clears any work or temporary tables that are created by the connection

• Kills all global cursors owned by the connection

• Closes any open SQL-XML handles that are open

• Deletes any open SQL-XML related work tables

• Closes all system tables

• Closes all user tables

• Drops all temporary objects

• Aborts open transactions

• Defects from a distributed transaction when enlisted

• Decrements the reference count for users in current database which releases shared database locks

• Frees acquired locks

• Releases any acquired handles

• Resets all SET options to the default values

• Resets the @@rowcount value

• Resets the @@identity value

• Resets any session level trace options using dbcc traceon()

• Resets CONTEXT_INFO to NULL in SQL Server 2005 and newer _{[ not part of the original article ]}

sp_reset_connection will NOT reset:

• Security context, which is why connection pooling matches connections based on the exact connection string

• Application roles entered using sp_setapprole, since application roles could not be reverted at all prior to SQL Server 2005. Starting in SQL Server 2005, app roles can be reverted, but only with additional information that is not part of the session. Before closing the connection, application roles need to be manually reverted via sp_unsetapprole using a "cookie" value that is captured when sp_setapprole is executed.

Note: I am including the list here as I do not want it to be lost in the ever transient web.