In the overview below we list the important items that you should consider when choosing server hardware.
|Processors||Windows 2012 requires 64-bit processors. Although 32-bit applications will run, the operating systems only comes in a 64-bit edition.If CPU is the limiting resource in the system, a core with 2x frequency typically provides a greater performance improvement than two cores with 1x frequency. Multiple cores are not expected to provide a perfect linear scaling, and the scaling factor can be even less if hyperthreading is enabled because hyperthreading relies on sharing resources of the same physical core.|
|Cache||Opt for processors with large L2 or L3 caches. The larger caches generally provide better performance, and often have a higher performance impact than raw CPU frequency.|
|Memory (RAM) and paging storage||Increase the RAM to match your memory needs.When your computer runs low on memory and it needs more immediately, modern operating systems use the page file on a hard disk to supplement system RAM. Paging will degrade the overall system performance.
You can optimize paging by using the following guidelines:
· Place the page file on a drive that is not fault-tolerant. Note that, if the disk fails, a system crash is likely to occur. If you place the page file on a fault-tolerant drive, remember that fault-tolerant systems are often slower to write data because they write data to multiple locations.
· Use multiple disks or a disk array if you need additional disk bandwidth for paging. Do not place multiple page files on different partitions of the same physical disk drive.
|Peripheral bus||With Windows Server 2012, it is highly recommended that the primary storage and network interfaces are PCI Express (PCIe). To avoid bus speed limitations, use PCIe x8 or higher slots for 10 Gigabit Ethernet adapters.|
|Disks||Choose disks with higher rotational speeds to reduce random request service times (~2 ms on average when you compare 7,200- and 15,000-RPM drives) and to increase sequential request bandwidth. Keep in mind that there are cost, power, and other considerations associated with disks that have high rotational speeds.Also, remember that consolidating small drives into fewer high-capacity drives most of the time will reduce overall storage performance. Fewer spindles mean reduced request service concurrency.|
The table below lists recommendations for network and storage adapters for high-performance servers which will help prevent your networking or storage hardware from being a bottleneck.
|WHQL certified||The adapter has passed the Windows® Hardware Quality Labs (WHQL) certification test suite.|
|64-bit capability||Adapters that are 64-bit-capable can perform direct memory access (DMA) operations to and from high physical memory locations (greater than 4 GB). If the driver does not support DMA greater than 4 GB, the system double-buffers the I/O to a physical address space of less than 4 GB.|
|Copper and fiber (glass) adapters||Copper adapters generally have the same performance as their fiber counterparts, and both copper and fiber are available on some Fibre Channel adapters. Certain environments are better suited to copper adapters, whereas other environments are better suited to fiber adapters.|
|Dual- or quad-port adapters||Multiport adapters are useful for servers that have a limited number of PCI slots.To address SCSI limitations on the number of disks that can be connected to a SCSI bus, some adapters provide two or four SCSI buses on a single adapter card. Fibre Channel disks generally have no limits to the number of disks that are connected to an adapter unless they are hidden behind a SCSI interface.
Serial Attached SCSI (SAS) and Serial ATA (SATA) adapters also have a limited number of connections because of the serial nature of the protocols, but you can attach more disks by using switches.
Network adapters have this feature for load-balancing or failover scenarios. Using two single-port network adapters usually yields better performance than using a single dual-port network adapter for the same workload.
PCI bus limitation can be a major factor in limiting performance for multiport adapters. Therefore, it is important to consider placing them in a high-performing PCIe slot that provides enough bandwidth.
|Interrupt moderation||Some adapters can moderate how frequently they interrupt the host processors to indicate activity or its completion. Moderating interrupts can often result in reduced CPU load on the host, but unless interrupt moderation is performed intelligently, the CPU savings might increase latency.|
|Receive Side Scaling (RSS) support||RSS is a technology that enables packet receive-processing to scale with the number of available computer processors. Particularly important with faster Ethernet (10 GB or more).|
|Offload capability and other advanced features such as message-signaled interrupt (MSI)-X||Offload-capable adapters offer CPU savings that yield improved performance. For more information, see Choosing a Network Adapter later in this guide.|
|Dynamic interrupt and deferred procedure call (DPC) redirection||Windows Server 2012 has functionality that enables PCIe storage adapters to dynamically redirect interrupts and DPCs. This capability, originally called “NUMA I/O,” can help any multiprocessor system by improving workload partitioning, cache hit rates, and on-board hardware interconnect usage for I/O-intensive workloads.|
Although we mostly focus on performance of Windows 2012 in this series, it is important to recognize the increasing importance of energy efficiency in data center environments. Below we provide some guidelines for power characteristics and capabilities of server components.
|Processors||The energy consumption of processors is affect by its frequency, operating voltage, cache size, and process technology. Processors have a thermal design point (TDP) rating that gives a basic indication of energy consumption relative to other models. In general, opt for the lowest TDP processor that will meet your performance goals. Newer processors are typically more energy efficient, and may expose more power states for Windows power management.|
|Memory (RAM)||Memory accounts for an increasing fraction of the total system power. Many factors affect the energy consumption of a memory DIMM, such as memory technology, error correction code (ECC), bus frequency, capacity, density, and number of ranks. It is best to compare expected power ratings before purchasing large quantities of memory. Low-power memory is now available, but you should consider the performance and cost trade-offs.|
|Disks||Higher RPM means increased energy consumption. Also, 2.5-inch drives generally require less power than 3.5-inch drives.|
|Network and storage adapters||Some adapters can decrease energy consumption during idle periods. This is an important consideration for 10 Gb networking adapters and high-bandwidth (4-8 Gb) storage links. Such devices can consume significant amounts of energy.|
|Power supplies||Increasing power supply efficiency is a great way to reduce energy consumption without affecting performance. High-efficiency power supplies can save many kilowatt-hours per year, per server.|
|Fans||Fans, like power supplies, are an area where you can reduce energy consumption without affecting system performance. Variable-speed fans can reduce RPM as the system load decreases, eliminating otherwise unnecessary energy consumption.|
|USB devices||Windows Server 2012 enables selective suspend for USB devices by default. However, a poorly written device driver can still disrupt system energy efficiency by a sizeable margin. To avoid potential issues, disconnect USB devices, disable them in the BIOS, or choose servers that do not require USB devices.|
|Remotely managed power strips||Power strips are not an integral part of server hardware, but they can make a large difference in the data center. Measurements show that volume servers that are plugged in, but have been ostensibly powered off, may still require up to 30 watts of power. To avoid wasting electricity, you can deploy a remotely managed power strip for each rack of servers to programmatically disconnect power from specific servers.|
In our next article, we’ll get more in the power and performance considerations related to Windows Server 2012.