Shulou(Shulou.com)06/01 Report--

This article shows you the analysis of the use of Windows 2003's own performance monitoring tools, which is concise and easy to understand, which will definitely brighten your eyes. I hope you can get something through the detailed introduction of this article.

A few days ago, after a standby library accumulated REDO for many days, the whole system became extremely slow during REDO APPLY. Through the Windows 2003 performance monitoring tool, it was found that the hard disk IO appeared a bottleneck. The following is a list of metrics for the collected Windows 2003 performance monitoring tools.

How the tool is executed: manage-> performance

Related performance indicators:

1) Memory-> Page Reads/sec is too high and lasts longer than 5.

Too much page swapping uses a lot of hard disk space, and it is easy to confuse insufficient memory for page swapping with the disk bottle diameter that causes page swapping. This requires us to track the following disk usage counters and memory counters when studying the reasons for page swapping that is not obviously out of memory:

Physical Disk->% Disk Time

Physical Disk-> Avg.Disk Queue Length

Memory-> Page Reads/sec

Meaning:

Page Reads/sec is the number of times the disk was read to parse hard page errors. It shows the number of read operations, regardless of the number of pages for each operation. A hard page error occurs when a process references a page of virtual memory that is located outside the working set or somewhere else in physical memory, and the page must be retrieved from disk. This counter is the main indicator that causes system-wide delays. It contains errors that read operations to satisfy file system caching (usually requested by the application) and non-cache mapped memory files. Compare the value of memory\\ PagesReads/sec with the value of memory\\ PagesInput/sec to determine the average number of pages read by each operation.

Popular meaning:

The hard failure of the page, a subset of Page/sec, is the number of times it is read from the hard disk to resolve hard errors.

Analysis:

The threshold is > 5, and the lower the better. A large number indicates a disk read rather than a cache read. A value that persists greater than 5, indicating that there are more page-missing interruptions (page fault) in memory read requests.

Indicates that the Working Set of the process is no longer enough, so the hard disk is used for virtual memory. Here is the number of reads, do not care about the number of pages read, a larger value indicates that there is a memory bottleneck.

A) if Memory-> Page Reads/sec is low, but Physical Disk->% Disk Time and Physical Disk-> Avg. The Disk Queue Length counter is high, indicating that the disk has a bottleneck.

B) if with Physical Disk-> Avg. The increase in Disk Queue Length, but not the decrease in Memory-> Page Reads/sec, indicates a memory bottleneck.

2) related performance indicators of hard disk

The Avg.Disk sec/Transfer counter reflects the time it takes the disk to complete the request. A high value indicates that the disk controller keeps retrying the disk due to failure. These failures increase the average disk transfer time. For most disks, the average transfer time for higher disks is more than 0.3 seconds.

An Avg.Disk Bytes/Transfer value greater than 20 KB indicates that the disk drive is usually running well; if the application is accessing the disk, it results in a lower value. For example, applications that randomly access the disk increase the average Disk sec/Transfer time because random transfers require increased search time.

Generally speaking, Disk Transfer/sec defines that a value less than 15ms is the best, a value between 15-30ms is good, and a value between 30-60ms is acceptable. If you exceed 60ms, you need to consider changing the hard disk or the RAID way to replace the hard disk.

If one of the following counters has a value greater than 20 milliseconds, the disk is overloaded: Avg.Disk sec/Read,avg.Disk sec/ Write,Avg.Disk sec/Transfer.

If only% Disk Time is large and all other values are moderate in the Physical Disk counter, the hard disk may be the bottleneck. If several values are relatively large and the value continues to exceed 80%, it may be a memory leak.

3) determine workload balance

To balance the load on a network server, you need to know how busy the server disk drives are. Use the Physical Disk->% Disk Time counter, which displays the percentage of drive activity time. If% Disk Time is high (more than 90%), check the Physical Disk-> Current Disk Queue Length counter to see the number of system requests waiting for disk access. The number of waiting Icano requests should be kept at no more than 1.5 to 2 times the number of spindles that make up the physical disk. Most disks have only one axis, but redundant Array of Independent disks (RAID) devices typically have multiple axes. The hardware RAID device appears as a physical disk in the system Monitor. Multiple RAID devices created by the software appear as multiple instances in the system Monitor.

You can use the values of Current Disk Queue Length and% Disk Time counters to detect bottlenecks in the disk subsystem. If the values of Current Disk Queue Length and% Disk Time counters are always high, consider the following:

Use a faster disk drive.

Move some files to another disk or server.

If you are using a RAID array, add disks to the array.

If you use a RAID device, the% Disk Time counter indicates a value greater than 100% (this occurs because the processor allows the operating system to use overlapping Ibank O, and the disk performance counter uses a counter with a precision of 100 nanoseconds to measure disk time, which is displayed at a sampling rate. The time after sampling may exceed 100%. For example, there are 10 requests within 2 milliseconds, and the sampling time is 10 milliseconds. If it is a raid disk, because the operating system can read and write to as much as the disk, resulting in overlapping Imax O, which makes the% disk time greater than 100%. ). If this occurs, use the PhysicalDisk: Avg.Disk Queue Length counter to determine the average number of system requests waiting for disk access.

Applications or systems on which Icano depends may keep the disk active.

Monitoring the Memory: Page Faults/sec counter ensures that disk activity is not caused by paging. In Windows, the reasons for page change include:

The configuration process takes up too much memory.

File system activity.

If you have multiple logical partitions on the same hard disk, use Logical Disk counters instead of Physical Disk counters. Looking at the logical disk counter helps determine which files are accessed frequently. When you find that there is a lot of read / write activity on the disk, look at the read-write counter to determine the type of disk activity that increases the load on each logical volume, for example, Logical Disk: Disk Write Bytes/sec.

Use skills:

1) the function of log tracking can be enabled to record the changes of various parameters in the log file for later analysis.

2) if there are too many indicators to see clearly, you can customize the counter to increase the counting index; the color, ratio, thickness of the line, etc., in order to distinguish the curve you are most concerned about.

The above is the analysis of the use of Windows 2003's built-in performance monitoring tools. Have you learned any knowledge or skills? If you want to learn more skills or enrich your knowledge reserve, you are welcome to follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.