The Dell EMC SC Series is a block storage system aimed at the mid-tier enterprise market and was formerly called the Compellent family. The models include the SC9000, SC7020, SC5020 (and F models as all flash) and the SCv3000 as a family targeted at the lowest cost entry market. The architecture of the SC systems is that data can be automatically stored, moved, exported, or imported based on the intelligence in the SC system. The advanced features of the SC system are centered on managing and moving data to optimize the storage.
The Dell EMC Storage Center is a dual controller design with redundant components that allow for failover in the case of failure of a controller and continued operation in the case of failure of other components. The dual controller system uses communication through switches to detect failure in the paired controller and handle the failover. Physical disks are virtualized with blocks to be stored or retrieved managed across the disks in virtual volumes. Using the indirection with the block locations not tied to specific disks prior to storing allows for the features such as automated tiering, use of all physical resources, and redistribution of data to avoid hotspots to be implemented. The virtualized architecture with blocks being handled independently also gives flexibility in handling of the RAID protection and the ability to do wide striping allowing for more simultaneous disk access.
Dell EMC supports multiple types of SSDs as separate tiers of storage with migration between the tiers handled by the standard Data Progression software. 3D NAND flash devices are supported with up to 15TB capacity.
Common management is used across all SC models with Unisphere for SC, which is an HTML5 web GUI.
The cache management allows for multiple threads to access cache with read-ahead algorithm for read caching of data in system memory and a separate write cache mirrored across the controllers. The write cache is battery backed to maintain data in the case of a failure. I/O operations are monitored and optimized for performance by merging contiguous block requests into a single request to minimize physical disk revolutions, sorting I/O request to minimize seek operations, and deep-tagged command queuing for multiple outstanding I/O requests to disks.Download the full Product Brief now!