在Join中使用DateDiff

Question

我尝试查询我们的登录日志，以确定自2017年7月1日起多个用户在30秒内登录到同一IP地址的情况。我在自我加入时使用了datediff但是给出了登录表有大约220万条记录，查询性能很糟糕。任何人都可以建议比下面的SQL更有效的查询。

SELECT TOP 1000 a.[id]
      ,A.[userId]
      ,AName.Firstname + ' ' + AName.LastName as name
      ,A.[username]
      ,A.[site]
      ,A.[loginTime]
      ,A.[ip]
      ,A.[agent]
      ,B.userid
      ,bname.Firstname + ' ' + bname.Lastname
      ,B.username 
      ,B.ip
      ,B.loginTime

  FROM LoginLog as A 
  Inner Join Loginlog As B on a.ip = B.ip and A.userId <> B.userId and DATEDIFF(SECOND, A.loginTime, B.loginTime) < 30 and DATEDIFF(SECOND, A.loginTime, B.loginTime) > 0 
  inner join Employees as AName on A.userId = AName.EmployeeNumber  
  outer join Employees as BName on B.userId = BName.EmployeeNumber 
  where a.userId in (*employee number list*) 
    and a.loginTime > '7/1/2017' 
  order by a.loginTime