| Now that you know how to monitor your system, let's look at how you can tune the operating system and hardware performance. The areas I'll examine are the following: Monitoring and Tuning Memory Usage Memory is often the source of performance problems, and you should always rule out memory problems before examining other areas of the system. Systems use both physical and virtual memory. To rule out memory problems with a system, you should configure application performance, memory usage, and data throughput settings, and then monitor the server's memory usage to check for problems. Setting Application Performance and Memory Usage Application performance and memory usage settings determine how system resources are allocated. In most cases, you want to give the operating system and background applications the lion's share of resources. This is especially true for Active Directory, file, print, and network and communications servers. On the other hand, for application, database and streaming media servers, you'll want to give the programs the server is running the most resources. To check these settings, follow these steps: -
Start the System utility from the Control Panel. -
Access the Advanced tab in the System utility, and then display the Performance Options dialog box by clicking Settings on the Performance panel. Choose the Advanced tab. -
The Processor Scheduling panel controls allocation of CPU time. To give more CPU time to the operating system and background services, select Background Services. Otherwise, select Programs. -
The Memory Usage panel controls allocation of memory. To give more memory to the operating system and background services, select System Cache. Otherwise, select Programs. -
Click OK. Setting Data Throughput Data throughput settings control how well a server responds to user requests , file handles, and client connections. You can optimize data throughput settings in one of four ways: -
Minimize Memory Used Optimizes the server to serve a small number of users. This means very little system memory is reserved for user requests, file handles, and client connections. This allows the server to reserve memory for other purposes but doesn't necessarily reduce the size of the system cache or reserved memory. (You'll experience poor responsiveness and performance if there are a lot of user requests, file handles, and client connections.) -
Balance Optimizes the server for mixed usage environments where the server has multiple roles that include file and printer sharing as well as other tasks . This results in average responsiveness to requests, file handles, and client connections. -
Maximize Data Throughput For File Sharing Optimizes the server to handle file and print services. This means the server will dedicate as many resources as possible to handling user requests, file handles, and client connections, which improves responsiveness and can also improve performance for user, file, and client actions. -
Maximize Data Throughput For Network Applications Optimizes the server memory for distributed applications that manage their own memory cache, such as SQL Server and IIS. This reduces the size of the system cache and allows more memory to be allocated to applications. To configure data throughput, complete the following steps: -
Access Network Connections in Control Panel. -
Right-click Local Area Connection, and then select Properties. This displays the Properties dialog box. Servers with multiple network interface cards will have multiple network connections shown in Network Connections. You should optimize each of these connections appropriately. -
Select File And Printer Sharing For Microsoft Networks, and then click Properties. -
On the Server Optimization tab, select the appropriate optimization setting. Click OK. -
You'll need to reboot the server for these changes to take effect. Checking Memory, Caching, and Virtual Memory Usage Now that you've optimized the system, you can determine how the system is using memory and check for problems. Table 3-1 provides an overview of counters that you'll want to track to uncover memory, caching, and virtual memory (paging) bottlenecks. The table is organized by issue category. Table 3-1. Uncovering Memory-Related Bottlenecks | Issue | Counters to Track | Details | | Physical and virtual memory usage | Memory\ Available Kbytes Memory\ Committed Bytes | Memory\Available Kbytes is the amount of physical memory available to processes running on the server. Memory\Committed Bytes is the amount of committed virtual memory. If the server has very little available memory, you might need to add memory to the system. In general, you want the available memory to be no less than 5 percent of the total physical memory on the server. If the server has a high ratio of committed bytes to total physical memory on the system, you might need to add memory as well. In general, you want the committed bytes value to be no more than 75 percent of the total physical memory. | | Memory page faults | Memory\Page Faults/sec Memory\Pages Input/sec Memory\Page Reads/sec | A page fault occurs when a process requests a page in memory and the system can't find it at the requested location. If the requested page is elsewhere in memory, the fault is called a soft page fault . If the requested page must be retrieved from disk, the fault is called a hard page fault . Most processors can handle large numbers of soft faults. Hard faults, however, can cause significant delays. Page Faults/sec is the overall rate at which the processor handles all types of page faults. Pages Input/sec is the total number of pages read from disk to resolve hard page faults. Page Reads/sec is the total disk reads needed to resolve hard page faults. Pages Input/sec will be greater than or equal to Page Reads/sec and can give you a good idea of your hard page fault rate. If there are a high number of hard page faults, you may need to increase the amount of memory or reduce the cache size on the server. | | Memory paging | Memory\Pool Paged Bytes, Memory\Pool Nonpaged Bytes | These counters track the number of bytes in the page and nonpaged pool. The paged pool is an area of system memory for objects that can be written to disk when they aren't used. The nonpaged pool is an area of system memory for objects that can't be written to disk. If the size of the page pool is large relative to the total amount of physical memory on the system, you might need to add memory to the system. If the size of the nonpaged pool is large relative to the total amount of virtual memory allocated to the server, you might want to increase the virtual memory size. | Monitoring and Tuning Processor Usage The CPU does the actual processing of information on your server. As you examine a server's performance, you should focus on the CPU after memory bottlenecks have been eliminated. If the server's processors are the performance bottleneck, adding memory, drives , or network connections won't overcome the problem. Instead, you might need to upgrade the processors to faster clock speeds or add processors to increase the server's upper capacity. You could also move processor- intensive applications, such as SQL Server, to another server. Before you make a decision to upgrade CPUs or add CPUs, you should rule out problems with memory and caching. If signs still point to a processor problem, you should monitor the performance counters discussed in Table 3-2. Be sure to monitor these counters for each CPU installed on the server. Table 3-2. Uncovering Processor-Related Bottlenecks | Issue | Counters to Track | Details | | Thread queuing | System\Processor Queue Length | This counter displays the number of threads waiting to be executed. These threads are queued in an area shared by all processors on the system. If this counter has a sustained value of two or more threads, you'll need to upgrade or add processors. | | CPU usage | Processor\ %Processor Time | This counter displays the percentage of time the selected CPU is executing a nonidle thread. You should track this counter separately for all processor instances on the server. If the % Processor Time values are high while the network interface and disk I/O throughput rates are relatively low, you'll need to upgrade or add processors. | Monitoring and Tuning Disk I/O With today's high-speed disks, the disk throughput rate is rarely the cause of a bottleneck. That said, however, accessing memory is much faster than accessing disks. So, if the server has to do a lot of disk reads and writes , the server's overall performance can be degraded. To reduce the amount of disk I/O, you want the server to manage memory very efficiently and page to disk only when necessary. You monitor and tune memory usage as discussed previously in the "Monitoring and Tuning Memory Usage" section of this chapter. Beyond the memory tuning discussion, you can monitor some counters to gauge disk I/O activity. Specifically, you should monitor the counters discussed in Table 3-3. Table 3-3. Uncovering Drive-Related Bottlenecks | Issue | Counters to Track | Details | | Overall drive performance | PhysicalDisk\% Disk Time in conjunction with Processor\% Processor Time and Network Interface Connection\Bytes Total/sec | If the % Disk Time value is high and the processor and network connection values aren't high, the system's hard disk drives might be creating a bottleneck. Be sure to monitor % Disk Time for all hard disk drives on the server. | | Disk I/O | PhysicalDisk\Disk Writes /sec, Physical Disk\ Disk Reads/sec, Physical Disk\Avg. Disk Write Queue Length, Physical Disk\Avg. Disk Read Queue Length, Physical Disk\Current Disk Queue Length | The number of writes and reads per second tell you how much disk I/O activity there is. The write and read queue lengths tell you how many write or read requests are waiting to be processed . In general, you want there to be very few waiting requests. Keep in mind that the request delays are proportional to the length of the queues minus the number of drives in a redundant array of independent disks (RAID) set. | Monitoring and Tuning Network Bandwidth and Connectivity No other factor weighs more in the way a user perceives your server's performance than the network that connects your server to the user's computer. The delay, or latency, between when a request is made and the time it's received can make all the difference. If there's a high degree of latency, it doesn't matter if you have the fastest server on the planet. The user experiences a delay and perceives that your servers are slow. Generally speaking, the latency the user experiences is beyond your control. It's a function of the type of connection the user has and the route the request takes to your server. The total capacity of your server to handle requests and the amount of bandwidth available to your servers are factors under your control, however. Network bandwidth availability is a function of your organization's network infrastructure. Network capacity is a function of the network cards and interfaces configured on the servers. The capacity of your network card can be a limiting factor in some instances. Most servers use 10/100 network cards, which can be configured in many ways. Someone might have configured a card for 10 Mbps or the card might be configured for half duplex instead of full duplex. If you suspect a capacity problem with a network card, you should always check the configuration. To determine the throughput and current activity on a server's network cards, you can check the following counters: If the total bytes per second value is more than 50 percent of the total capacity under average load conditions, your server might have problems under peak load conditions. You might want to ensure that operations that take a lot of network bandwidth, such as network backups , are performed on a separate interface card. Keep in mind that you should compare these values in conjunction with PhysicalDisk\% Disk Time and Processor\% Processor Time. If the disk time and processor time values are low but the network values are very high, there might be a capacity problem. Solve the problem by optimizing the network card settings or by adding an additional network card. Remember, planning is everything ”it isn't always as simply as inserting a card and plugging it into the network.  |
![Microsoft Windows Server 2003 Administrator[ap]s Pocket Consultant](/icons/blank_book.jpg "Microsoft Windows Server 2003 Administrator[ap]s Pocket Consultant"){kind=icon}