Resilient File System Overview
Published: February 29, 2012
Updated: February 29, 2012
Applies To: Windows Server 2012
This topic describes the new Resilient File System (ReFS) in Windows Server® 2012, the practical uses for the new file system, and the most significant new functionality.
Windows customers want a cost-effective platform that maximizes data availability, scales efficiently to very large data sets across diverse workloads, and guarantees data integrity via resiliency to corruption (regardless of software or hardware failures). ReFS is a new file system that targets these customer needs while providing a foundation for significant future innovations. By utilizing an integrated storage stack comprising ReFS and the new Storage Spaces feature in Windows Server 2012, customers can now deploy the most cost-effective platform for available and scalable data access using commodity storage.
Storage Spaces protects data from partial and complete disk failures by maintaining copies on multiple disks. ReFS interfaces with Storage Spaces to automatically repair the corruption. For more information on Storage Spaces, see File and Storage Services Overview.
The key features of ReFS include:
Maintains the highest levels system availability and reliability possible under the assumption that underlying storage may be inherently unreliable.
Provides a full end-to-end resilient architecture when used in conjunction with Storage Spaces so that these two features magnify the capabilities and reliability of the other when used together.
Maintains compatibility with the NTFS features that are widely adopted and successful while replacing features that provide limited value.
In addition, there is a new Integrity.exe command line tool to manage the integrity and disk scrubbing policies.
ReFS provides functionality that helps customers store and protect data, regardless of the reliability of the underlying hardware and software stack. This minimizes the cost of storage and reduces capital expenditures for businesses. Some practical applications include:
General-purpose file server. Customer deploys a file server attached to a JBOD storage configuration with Serial ATA (SATA) or Serially Attached SCSI (SAS) drives.
Consolidated remote application data storage. Customer deploys a scale-out, two-node file server cluster with Storage Spaces, in which the cluster uses a shared JBOD storage configuration with SATA or SAS drives.
The significant functionality included with ReFS is described below:
Integrity. ReFS stores data in a way that protects it from many of the common errors that can normally cause data loss. When ReFS is used in conjunction with a mirrored Storage Space, detected corruption – both metadata and user data, when integrity streams are enabled - can be automatically repaired using the alternate copy provided by Storage Spaces. In the event of a system error, ReFS quickly recovers from the error with no loss of user data.
Availability. ReFS prioritizes the availability of data. Historically, file systems were often susceptible to data corruption that would require the system to be taken offline for repair. With ReFS, if corruption occurs, the repair process is both localized to the area of corruption and performed online, requiring no volume down-time. Although rare, if a volume does become corrupted, or you choose not to use mirrored Storage Spaces, ReFS implements salvage, a feature that removes the corrupt data from the namespace on a live volume and good data is not adversely affected by non-repairable corrupt data. In addition, there is no chkdsk with ReFS.
Scalability. As the amount and size of data that is stored on computers continues to rapidly increase, ReFS is designed to work well with extremely large data sets, petabytes and larger, without performance impact. While practical concerns surrounding system configurations (such as the amount of memory), limits by various system components and the time taken to populate data sets or backup times may define practical limitations. The ReFS on-disk format is designed to support volume sizes up to 2^78 bytes using 16KB cluster sizes while Windows stack addressing allows 2^64 bytes. This format also supports 2^64-1 byte file sizes, 2^64 files in a directory, and the same number of directories in a volume.
Proactive Error Identification. The integrity capabilities of ReFS are leveraged by a data integrity scanner, which is known as a scrubber. A scrubber periodically scans the volume, attempting to identify latent corruption, and then proactively triggering a repair of that corrupt data.
The following list provides additional resources on the web about related technologies in Windows Server 2012.