Hi,
We have sql server 2008r2 active-passive cluster, running on windows 2003. It has 16CPUs, 128GB of RAM, and all databases (including tempdb) are placed on the san, configured with 160 spindles. Tempdb is spread across 16 files.
Recently we started getting these warnings about slowness on tempdb files (NEVER user databases):
09/17/2012 12:04:55,spid4s,Unknown,SQL Server has encountered 884 occurrence(s) of I/O requests taking longer than 15 seconds to complete on file [D:\MSSQL10\MSSQL\Data\tempdev15.ndf] in database [tempdb] (2). The OS file handle is 0x0000000000000AC4. The offset of the latest long I/O is: 0x0000007e500000
It is ALWAYS spid 4 (so not some user spid). When I pull info from dynamic management views about sql that ran approximately at that time, I see nothing with a lot of IOs:
select top 10 (total_logical_reads/execution_count) AS avg_logical_reads,(total_logical_writes/execution_count) AS avg_logical_writes,
(total_physical_reads/execution_count) AS avg_phys_reads,
execution_count,
(select substring(text, statement_start_offset/2 + 1,
(case when statement_end_offset = -1
then len(convert(nvarchar(max),text)) * 2
else statement_end_offset
end - statement_start_offset)/2)
from sys.dm_exec_sql_text(sql_handle)) as query_text,
last_execution_time,
last_elapsed_time/1000000 as elapsed_seconds
from sys.dm_exec_query_stats
where last_execution_time between <start time> and <end time>
order by (total_logical_reads + total_logical_writes) desc
Does anybody have any idea how to troubleshoot this (other than running the trace and hoping that I will catch it)? What engine system process uses spid 4?
Thanks
Pedja