Performance Tuning for File Server Workload (SPECsfs2008)
SPECsfs2008 is a file server benchmark suite from Standard Performance Evaluation Corporation that measures file server throughput and response time, providing a standardized method for comparing performance across different vendor platforms. SPECsfs2008 results summarize the server's capabilities with respect to the number of operations that can be handled per second, and the overall latency of the operations.
To ensure accurate results, you should format the data volumes between tests to flush and clean up the working set. For improved performance and scalability, we recommend that you partition client data over multiple data volumes. The networking, storage, and interrupt affinity sections of this guide contain additional tuning information that might apply to specific hardware.
Registry-Tuning Parameters for NFS File Servers
You can tune the following registry parameters previously described in the Tuning Parameters for NFS File Servers section to enhance the performance of NFS servers:
You can improve the performance of Active Directory®, especially in large environments, by following these tuning steps:
Use an appropriate amount of RAM.
Active Directory uses RAM to cache as much of the directory database as possible. This reduces disk access and improves performance. The Active Directory database can grow, but the amount of data that can be cached is limited by the virtual address space and how much physical RAM is on the server.
To determine whether more RAM is needed for the server, monitor the percentage of Active Directory operations that are being satisfied from the cache by using the Reliability and Performance Monitor. Examine the lsass.exe instance (for Active Directory Domain Services) or the Directory instance (for Active Directory Lightweight Directory Services) of the Database\Database Cache % Hit performance counter. A low value indicates that many operations are not being satisfied from the cache. Adding more RAM might improve the cache hit rate and the performance of Active Directory. You should examine the counter after Active Directory has been running for several hours under a typical workload. The cache starts out empty when the Active Directory service is restarted or when the computer is rebooted, so the initial hit rate is low.
Use a good disk I/O subsystem.
Ideally, the server is equipped with sufficient RAM to cache the “hot” parts (that is, the most frequently parts) of the database entirely in memory. However, the on-disk database must be accessed to initially populate the memory cache, when it accesses uncached parts of the database, and when it writes updates to the directory. Therefore, an appropriate selection for storage is important to Active Directory performance.
We recommend that the Active Directory database folder be located on a physical volume that is separate from the Active Directory log file folder. Both folders should be on a physical volume that is separate from the operating system volume. The use of drives that support command queuing, especially Serial Attached SCSI or SCSI, might also improve performance.
Considerations for Read-Heavy Scenarios
The typical directory workload consists of more query operations than update operations. Active Directory is optimized for such a workload. To obtain the maximum benefit, the most important performance tuning step is to make sure that the server has sufficient RAM to cache the most frequently used part of the database in memory. Query performance on a recently rebooted server, or after the Active Directory service is restarted, might initially be low until the cache is populated. Active Directory automatically populates the cache as queries search data in the directory.
Considerations for Write-Heavy Scenarios
Write-heavy scenarios do not benefit as much from the Active Directory cache. To guarantee the transactional durability of data that is written to the directory, Active Directory does not perform disk write caching It commits all write operations to the disk before it returns a successful completion status for an operation, unless there is an explicit request to not do this. Therefore, fast disk I/O is important to the performance of write operations to Active Directory. The following are hardware recommendations that might improve performance for these scenarios:
To determine whether disk I/O is a bottleneck, monitor the Physical Disk\Average Disk Queue Length counter for the volumes on which the Active Directory database and logs are located. A high queue length indicates a large amount of concurrent disk I/O activity. Choosing a storage system to improve write performance on those volumes might improve Active Directory performance.