今天在论坛看到一篇文章问到为什么SQL Server系统的SESSION ID是从1到50的，以前看过文章，这些Session是SQL Server为了运行系统活动比如（(lazy writer, ghost record cleanup, DTC commit/abort），所以会保留50 Session ID给SQL Server使用,用户会话从51开始。
在2005之前查询用户会话使用：
select * from sysprocesses wherespid < 50
但是在SQL Server 2005之后已经没有这个限制了，我在MSDN上找到了下面这篇文章：How It Works: System Sessions
Looking at a SQL Server error log it is formatted with the date, time and session identifier. Many of the identifiers contain the s following the spid value.
The s indicates that the session is a system session. Prior to SQL Server 2005 all system sessions were limited to session ids less than 50. SQL Server 2005 lifted that restriction. In order identify a session performing system actives (lazy writer, ghost record cleanup, DTC commit/abort, ...) the sessions are identified as system sessions.
Instead of the older "select * from sysprocesses where spid < 50" you should use "select * from sys.dm_exec_sessions where is_user_process = 0" to identify system processes.
根据上面的讲法系统Session ID有可能超过50，所以查询用户Session ID使用
select * from sys.dm_exec_sessionswhereis_user_process= 1
更多信息参考：http://msdn.microsoft.com/zh-cn/library/ms176013(v=sql.90).aspx