The Enterprise Resources Training Co.,Ltd.

Performance Counters
Performance object LogicaIDisk(*) Counter Avg. Disk Queue Length Description Shows the average numbers of both read and write requests that are queued for the selected disk during the sample interval. Threshold Avg. Value> 1.5 per spindle Always be aware of the physical architecture of your disk subsystem. If your drive consists of 10 physical disks in a redundant array of independent disks (RAID) 0+1 array, there are 5 usable physical disks in the array. Maximum Value >10 megabytes (MB)/sec Average Value < 128 A problem if server is paging.

Disk Bytes/sec

Memory

Available Mbytes

Free System Page Table Entries (PTE)

Shows the rate at which bytes are transferred to or from the disk during write or read operations. Shows the size, in bytes, of the nonpaged pool. Memory\Pool Nonpaged Bytes is calculated differently from how Process\Pool Nonpaged Bytes is. So, it might not be equal to Process (Total)\Pool Nonpaged Bytes. Shows the number of page table entries not currently in use by the system.

Pages/sec

Shows the rate at which pages are read from or written to the disk to resolve hard page faults. This counter is a primary indicator of the kinds of faults that cause system-wide delays. It is the sum of Memory\Pages, Input/sec, and Memory\Pages Output/sec. It is counted in number of pages so that it can be compared to other counts of pages, such as Memory\Page Faults/sec, without conversion. It

< 3,000. Normally, PTE should be between 80,000 and 140,000 and at least 15,000 if using a 3-gigabyte (GB) switch. Average Value < 5 More than 20 Pages/sec for a period of time is a definite cause for concern.

Performance object

Counter

Description includes pages retrieved to satisfy faults in the file system cache, usually requested by applications, and noncached mapped memory files. Indicates the rate at which bytes are sent and received over each network adapter, including framing characters. Network Interface\Bytes Received/sec is the sum of Network Interface\Bytes Received/sec and Network Interface\Bytes Sent/sec. Indicates the percentage of elapsed time that the process threads spent executing code in privileged mode. When a Microsoft Windows system service is called, the service often runs in privileged mode to gain access to system-private data. Such data is protected from access by threads executing in user mode. Calls to the system can be explicit or implicit, such as page faults or interrupts. Unlike some early operating systems, Windows uses process boundaries for subsystem protection in addition to the traditional protection of user and privileged modes. In addition to the privileged time in the process, some work done by the Windows operating system on behalf of the application might appear in other subsystem processes. Indicates the percentage of elapsed time that the processor spends to execute a non-idle thread. It is calculated by measuring the duration for which the idle thread is active in the sample interval and subtracting that time from the interval duration. Each processor has an idle thread that consumes cycles when no other threads are ready to run. This counter is the primary indicator of processor activity and displays the average percentage of busy time observed during the sample interval. This time is calculated by monitoring the time for which the service is inactive and then subtracting that value from 100%.

Threshold

Network Interface(*)

Bytes Total/sec

Average Value < 50% of network interface card (NIC) capacity.

Processor(_Total)

% Privileged Time

Average Value < 10 A high rate of privileged time might be attributable to a large number of interrupts generated by a failing device.

% Processor Time

Average Value > 80

Performance object Paging File(_Total)

Counter % Usage

Description Indicates the amount of Page File instance in use as a percentage.

Threshold > 70%. High percentage indicates shortfall of random access memory (RAM). > 90%. If threshold is reached, check disk queue counters to see whether there is really an input/output (I/O) issue. Average Value> 0.5 per spindle Always be aware of the physical architecture of your disk subsystem. If your drive consists of 10 physical disks in a RAID 0+I array, there are 5 usable physical disks in the array. > 1.5 GB

PhysicaIDisk(*)

% Disk Time

Shows the percentage of elapsed time that the selected disk drive is busy in servicing read or write requests.

Avg. Disk Queue Length

Shows the average numbers of both read and write requests that are queued for the selected disk during the sample interval.

Process(sqlservr)

Virtual Bytes

Processor(*)

% Privileged Time

Indicates the current size of the virtual address space the process is using in bytes. Use of virtual address space does not necessarily imply corresponding use of either disk or main memory pages. Virtual space is finite and the process can limit its ability to load libraries. Indicates the percentage of elapsed Average Value < 10 time that the process threads spent in executing code in privileged mode. When a Windows system service is called, the service often runs in privileged mode to gain access to the system-private data. Such data is protected from access by threads executing in user mode. Calls to the system can be explicit or implicit, such as page faults or interrupts. Unlike some early operating systems, the Windows operating system uses process boundaries for subsystem protection in addition to the traditional protection of user and privileged

Performance object

Counter

Description modes. Some work done by the Windows operating system on behalf of the application might appear in other subsystem processes in addition to the privileged time in the process. % Processor Time is the percentage of elapsed time that all process threads used the processor to execute instructions. An instruction is the basic unit ofexecution in a computer. A thread is the object that executes instructions. And a process is the object created when a program is run. The code executed to handle some hardware interrupts and trap conditions is included in this count. Indicates the current length of the server work queue for this central processing unit (CPU). A sustained queue length, greater than four, can indicate processor congestion. This is an instantaneous count and not an average over time. Shows the percentage of data requests served out of the buffer cache. Indicates the number of seconds a page will stay in the buffer pool without references. Shows the total number of free pages. Indicates the total wait time for log flushing in milliseconds. Indicates the ratio between cache hits and lookups. Indicates the current number of processes waiting for a workspace memory grant. Indicates the total amount of dynamic memory the server is willing to consume. Indicates the total amount of dynamic memory the server is currently consuming. Indicates the combined rate at which all processors on the computer are switched from one thread to another. Context switches occur when a rulU1ing thread voluntarily relinquishes the processor, is preempted by a

Threshold

% Processor Time

> 80%

Server WorkQueues(0)

Queue Length

>4

SQLServer: Buffer Manager

Buffer Cache Hit Ratio Page Life Expectancy Free Pages

Average Value <90% < 300 seconds

SQLServer: Database(*) SQLServer: Plan Cache _Total SQLServer: Memory Manager

Log Flush Wait Time Cache Hit Ratio Memory Grants Pending Target Server Memory (KB) Total Server Memory (KB)

Minimum value < 640 500 ms <90% Average Value = 0

Close to the size of physical memory. Close to the size of Target Server Memory. Avg. Value < 1,000 per processor

System

Context Switches/sec

Performance object

Counter

Description

Threshold

Processor Queue Length

higher-priority ready thread, or switches between user mode and privileged or kernel mode to use an executive or subsystem service. It is the sum of Thread\Context Switches/sec for all threads rulU1ing on all processors in the computer and is measured in number of switches. There are context switch counters on the System and Thread objects. This counter displays the difference between the values observed in the last two samples divided by the duration of the sample interval. > (10 x Number of Indicates the number of threads in the processor queue. Unlike disk counters, Processors) this counter shows only the ready threads and not the threads that are running. There is a single queue for processor time even on computers with multiple processors. Therefore, if a computer has multiple processors, you need to divide this value by the number of processors servicing the workload. A sustained processor queue of less than 10 threads per processor is normally acceptable, depending on the workload.