2008年11月7日 星期五

SQL 2005 重要效能計數物件(Performance Counter)及建議值

Object

Counter

Preferred Value

Description

Memory

Available Mbytes

> 100MB

 

Paging File

%Usage

< 70%

 

Process (sqlservr)

%Privileged Time

< 30% of %Processor Time (sqlservr)

 

Processor

%Privileged Time

< 30% of Total %Processor Time

 

PhysicalDisk

Avg. Disk Sec/Read

< 8ms

 

PhysicalDisk

Avg. Disk sec/Write

< 8ms (non cached) < 1ms (cached)

 

SQLServer:Access Methods

Forwarded Records/sec

< 10 per 100 Batch Requests/Sec

 

SQLServer:Access Methods

FreeSpace Scans/sec

<10 per 100 Batch Requests/Sec

 

SQLServer:Access Methods

Full Scans / sec

(Index Searches/sec)/(Full Scans/sec) > 1000

 

SQLServer:Access Methods

Page splits / sec compare with batch requests/sec  

SQLServer:Access Methods

Workfiles Created/Sec

< 20 per 100 Batch Requests/Sec

 

SQLServer:Access Methods

Worktables Created/Sec

< 20 per 100 Batch Requests/Sec

 

SQL Server:Buffer Manager

Buffer Cache hit ratio

> 90% , OLTP system 應大於 97%

Checkpoint pages/sec, Lay writes/sec, Page life expectancy

 

SQL Server:Buffer Manager

Checkpoint pages/sec

Page written do disk during the checkpoint process, free up SQL Server cache
Number of pages flushed to disk per second by a checkpoint or ther operation that required all dirty pages to be flushed
當下列情況發生時,代表 Memory pressure

High Checkpoint pages/sec

High lzay writes/sec

low page life expectancy (<300 seconds)
 

SQL Server:Buffer Manager

Free list stalls/sec

< 2

 

SQL Server:Buffer Manager

Lazy Writes/Sec

< 20

 

SQL Server:Buffer Manager

Page Life Expectancy
Time in seconds the data pages, on average, stay in SQL cache

> 300
小於300時,代表

  • Memory pressure
  • Memory constraints
  • Code SQL Server cache
  • Missing indexes, or poor query plans
 

SQLServer:Buffer Manager

Page lookups/sec

(Page lookups/sec) / (Batch Requests/sec) < 100

Number of requests to find a page in the buffer pool. When the ratio of page lookups to batch requests is much greater than 100, this is an indication that while query plans are looking up data in the buffer pool, these plans are inefficient. Identify queries with the highest amount of logical I/O's and tune them.

SQL Server:Locks

Lock Requests/sec

(Lock Request/sec)/(Batch Requests/sec) < 500

 

SQLServer:SQL Statistics

SQL Compilations/sec

< 10% of the number of Batch Requests/Sec

 

SQLServer:SQL Statistics

SQL Re-Compilations/sec

< 10% of the number of SQL Compilations

 

參考

https://blogs.msdn.com/jchiou/archive/2007/11/21/sql-2005-performance-counter.aspx

http://www.grumpyolddba.co.uk/monitoring/Performance%20Counter%20Guidance%20-%20SQL%20Server.htm

沒有留言:

Share with Facebook