Sql-server – Profiler or MDW

When you exhaust the upper bound of INT you will receive, for every new insert:

Msg 8115, Level 16, State 1, Line 1
Arithmetic overflow error converting IDENTITY to data type int.
Arithmetic overflow occurred.

Outside of the MDW use case

Converting to BIGINT is far safer (IMHO) than what some people do - go back to 0 and fill in the gaps, or go to -2 billion and just delay dealing with it until you use the same number of values one more time. If 2 billion rows isn't enough, neither is 4 billion. If you use compression changing the data type for values < 2 billion will actually save you some space.

However you will come across issues if this is part of a primary key constraint, referenced by foreign keys, has other constraints, etc. You'll need to perform some extra work in addition to just changing the data type.

For MDW specifically

The problem with the solution above, for MDW specifically, is that the data collector package (SSIS Packages/Data Collector/PerfCountersUpload) includes some logic that relies on the underlying data type to be INT. So, the collector jobs fail because of this mismatch.

This should buy you some time:

DBCC CHECKIDENT('snapshots.performance_counter_instances', RESEED, -2147483648);

And between now and when you start approaching 0 again (you should set up some kind of monitoring to check the max value and send some kind of alert when you get close), go in and clean up and make sure there are no positive values.

Then, when you start approaching 2 billion again, clean out all the negative values, and reseed.

Is this annoying? Absolutely. I think this might be one of several reasons why you don't see a very large MDW adoption.

You have to concentrate on Memory, Disk IO and CPU utilization to balance out your environment.

Best is to start collecting and dumping the results in a utility database and to give you a good baseline using sql agent on a scheduled basis.

Alternatively, you can use Performance data collector (applies to sql server 2008 and up)

MEMORY:

You can look into buffer pool usage (what databases are occupying more space in buffer pool ?)

From :Performance issues from wasted buffer pool memory

SELECT
    (CASE WHEN ([database_id] = 32767)
        THEN N'Resource Database'
        ELSE DB_NAME ([database_id]) END) AS [DatabaseName],
    COUNT (*) * 8 / 1024 AS [MBUsed],
    SUM (CAST ([free_space_in_bytes] AS BIGINT)) / (1024 * 1024) AS [MBEmpty]
FROM sys.dm_os_buffer_descriptors
GROUP BY [database_id];
GO

DISK IO:

In terms of Disk IO, you need to move away the mdf/ldf files having the most disk IO to a better/faster LUN/Array :

From : How to examine IO subsystem latencies from within SQL Server

SELECT
    [ReadLatency] =
        CASE WHEN [num_of_reads] = 0
            THEN 0 ELSE ([io_stall_read_ms] / [num_of_reads]) END,
    [WriteLatency] =
        CASE WHEN [num_of_writes] = 0
            THEN 0 ELSE ([io_stall_write_ms] / [num_of_writes]) END,
    [Latency] =
        CASE WHEN ([num_of_reads] = 0 AND [num_of_writes] = 0)
            THEN 0 ELSE ([io_stall] / ([num_of_reads] + [num_of_writes])) END,
    [AvgBPerRead] =
        CASE WHEN [num_of_reads] = 0
            THEN 0 ELSE ([num_of_bytes_read] / [num_of_reads]) END,
    [AvgBPerWrite] =
        CASE WHEN [num_of_writes] = 0
            THEN 0 ELSE ([num_of_bytes_written] / [num_of_writes]) END,
    [AvgBPerTransfer] =
        CASE WHEN ([num_of_reads] = 0 AND [num_of_writes] = 0)
            THEN 0 ELSE
                (([num_of_bytes_read] + [num_of_bytes_written]) /
                ([num_of_reads] + [num_of_writes])) END,
    LEFT ([mf].[physical_name], 2) AS [Drive],
    DB_NAME ([vfs].[database_id]) AS [DB],
    [mf].[physical_name]
FROM
    sys.dm_io_virtual_file_stats (NULL,NULL) AS [vfs]
JOIN sys.master_files AS [mf]
    ON [vfs].[database_id] = [mf].[database_id]
    AND [vfs].[file_id] = [mf].[file_id]
-- WHERE [vfs].[file_id] = 2 -- log files
-- ORDER BY [Latency] DESC
-- ORDER BY [ReadLatency] DESC
ORDER BY [WriteLatency] DESC;
GO

Also refer to Leveraging sys.dm_io_virtual_file_stats

CPU:

From : Brent's answer

SELECT total_worker_time/execution_count AS AvgCPU  
, total_worker_time AS TotalCPU
, total_elapsed_time/execution_count AS AvgDuration  
, total_elapsed_time AS TotalDuration  
, (total_logical_reads+total_physical_reads)/execution_count AS AvgReads 
, (total_logical_reads+total_physical_reads) AS TotalReads
, execution_count   
, SUBSTRING(st.TEXT, (qs.statement_start_offset/2)+1  
, ((CASE qs.statement_end_offset  WHEN -1 THEN datalength(st.TEXT)  
ELSE qs.statement_end_offset  
END - qs.statement_start_offset)/2) + 1) AS txt  
, query_plan
FROM sys.dm_exec_query_stats AS qs  
cross apply sys.dm_exec_sql_text(qs.sql_handle) AS st  
cross apply sys.dm_exec_query_plan (qs.plan_handle) AS qp 
ORDER BY 1 DESC

A gold mine of DMV related queries : SQL Server Diagnostic Information Queries for September 2014

For completeness, if you are using sql server 2012, you can use the system health reporting dashboard for Visualizing sp_server_diagnostics results.