DBCC SQLPERF(THREADS) returns a list of the currently running "threads" in the system, the information is equivalent to what is exposed via the sysprocesses table. The following query of sysprocesses provides an identical and better result set.
select [Spid] = spid, [Thread ID] = kpid, [Status] = convert(varchar(10), status), [LoginName] = convert(varchar(10), loginame), [IO] = physical_io, [CPU] = cpu, [MemUsage] = memusage from [master].[dbo].[sysprocesses] order by [spid] The result set: Spid Thread ID Status LoginName IO CPU MemUsage ------ ----------- ---------- ---------- ---------------------- ----------- ----------- 1 NULL sleeping NULL 0 0 0 2 NULL background NULL 0 0 5 3 NULL background NULL 0 0 -17 4 NULL background NULL 0 0 0 5 0 background sa 27 0 3 6 0 background sa 8 0 3 7 0 background sa 0 0 3 8 0 background sa 0 0 3 9 NULL sleeping NULL 17 0 0 10 0 background sa 0 0 3 11 0 background sa 0 0 3 12 0 background sa 0 0 3 13 0 background sa 0 0 3 14 0 background sa 0 0 3 15 0 background sa 0 0 3 16 0 background sa 0 0 3 51 0 sleeping GERTD00\AS 10 523 20 52 0 sleeping SDN\gertd 0 0 16 53 0 sleeping NT AUTHORI 4 0 4 54 0 sleeping NT AUTHORI 2 40 11 55 0 sleeping NT AUTHORI 227 380 46 56 1568 runnable NT AUTHORI 2 0 5 57 5532 runnable SDN\gertd 1 40 39 59 0 sleeping NT AUTHORI 0 0 6 60 0 sleeping NT AUTHORI 0 60 6 61 0 sleeping SDN\gertd 8 211 68 62 0 sleeping SDN\gertd 0 60 31 Description of the columns: | Column name | Description | | Spid | | | Thread ID | | | Status | | | LoginName | | | IO | | | CPU | | | MemUsage | |
The table structure returned, has the following shape: create table #threads ( [Spid] smallint not null, [Thread ID] int not null, [Status] nvarchar(10) not null, [Login Name] nvarchar(10) null, [IO] bigint not null, [CPU] int not null, [MemUsage] int not null ) To populate the table use: insert into #threads exec('dbcc sqlperf(threads) with tableresults, no_infomsgs') select * from #threads order by [Spid] NOTE: DBCC SQLPERF(THREADS, CLEAR) has no effect since the THREADS option is not based on cumulative statistical data. |