Caching technology has been around for decades alongside nearly all storage components/devices with the purpose of providing consistent and sustained performance to recently accessed data. Volatile memory technology, such as RAM, is typically employed for caching purposes due to its relatively low cost and industry standardization. While RAM is effective for most caching applications, it can prove inadequate in high write I/O infrastructures and all-flash storage deployments. Enlift cache is an innovative patented SSD-based caching technology from AMI StorTrends that greatly enhances both endurance and performance across the entire flash-based storage system. This is achieved by combining the advanced cache acceleration algorithms with state-of-the-art enterprise technology to ensure stability and consistent performance throughout the life of the array. Here we will take a look at the functionality of Enlift Cache, as well as the benefits it offers in flash-based storage production environments.
One of the two primary benefits from StorTrends’ Enlift Cache engine is the endurance it provides to the entire storage system. Among the 24 drives within the StorTrends all-flash SAN, 8 are leveraged as a ‘Hot Tier’ with no data compaction to deliver flash performance without overhead, 14 are utilized as a deduplicated and compressed ‘Cold Tier’, and the remaining 2 drives comprise the Enlift Cache layer. The Enlift Cache layer is utilized when migrating data between the two tiers, as well as all write I/O to data residing within the Cold Tier. While all drives within the system are enterprise-grade, the two drives that comprise the Enlift Cache layer have an unprecedented endurance rating of 10 Drive Writes Per Day (DWPD), meaning a single 400GB drive with this rating can withstand 4TB (400GBx10) of written day per day for 5 years before wearing out. The endurance rating of these drives is essential, since the Enlift Cache layer experiences the vast majority of write I/O on the system. By directing the majority of write I/O to a single high endurance, two-drive partition, the overall endurance of the system is increased significantly. This also largely contributes to the performance benefits from Enlift Cache.
The second primary benefit of the Enlift Cache layer is performance. When data (let’s say Block A) is initially written to the StorTrends all-flash SAN, Block A is immediately committed to the Hot Tier. Since this tier does not include data compaction technology, the user will realize instant flash performance without compaction overhead. If Block A is not accessed again within a fixed period of time set by the user, the block will be automatically migrated to the Cold Tier where it is deduplicated and compressed for additional space savings. Any write changes to Block A within the Cold Tier will instead be written directly to the Enlift Cache layer to avoid data compaction overhead and degradation of the Cold Tier SSDs. If Block A is continuously accessed, it will eventually be promoted back to the Hot Tier and freed from the Enlift Cache layer. While this example simply follows the cycle of a single block, one can infer the importance of the Enlift Cache when considering the +100,000 I/O accessing the system per second. In this sense, the Enlift Cache layer is a buffer between the Hot and Cold SSD Tiers that allows StorTrends All Flash SANs to deliver sustained immediate flash performance while simultaneously providing dense data compaction for unrivaled cost effectiveness.
See it For Yourself!
Overall, StorTrends’ Enlift Cache is a specialized and innovative technology that is instrumental in guaranteeing hardware endurance and system performance. Enlift Cache is what gives StorTrends All Flash SANs the edge over the competition in both SSD lifespan and extreme IOPS. If you would like to view a live demo of the Enlift Cache and learn how your IT environment can benefit from this technology, you can register here to register for an online product demo.