Client-side Flash Caching for Cloud Systems

Abstract

As the size of cloud systems and the number of hosted VMs rapidly grow, the scalability of shared VM storage systems becomes a serious issue. Client-side flash-based caching has the potential to improve the performance of cloud VM storage by employing flash storage available on the client-side of the storage system to exploit the locality inherent in VM IOs. However, because of the limited capacity and durability of flash storage, it is important to determine the proper size and configuration of the flash caches used in cloud systems. This paper provides answers to the key design questions of cloud flash caching based on dm-cache, a block-level caching solution customized for cloud environments, and a large amount of long-term traces collected from real-world public and private clouds. The study first validates that cloud workloads have good cacheability and dm-cache-based flash caching incurs low overhead with respect to commodity flash devices. It further reveals that write-back caching substantially outperforms write-through caching in typical cloud environments due to the reduction of server IO load. It also shows that there is a tradeoff on making a flash cache persistent across client restarts which saves hours of cache warm-up time but incurs considerable overhead from committing every metadata update persistently. Finally, to reduce the data loss risk from using write-back caching, the paper proposes a new cache-optimized RAID technique, which minimizes the RAID overhead by introducing redundancy of cache dirty data only, and shows to be significantly faster than traditional RAID and write-through caching.

An Empirical Evaluation of NVM Express SSD

“…For example, modern cloud environment takes advantage of SSD caching to improve the performance [4], [5], [6] by exploiting locality inherent in VM I/Os, thereby reducing the overall load of the storage system. The benefit of SSD caching becomes more significant as the speed of an SSD cache substantially outperforms HDD [4].…”

Flash Caching for Cloud Computing Systems

“…The write-invalidate policy performs poorly for write-intensive workloads. The write-through policy's performance is close to write-back when the write is submitted to the storage server asynchronously and the server's load is light [HAWS13]; otherwise, it can be substantially worse than the write-back policy [AZ14]. Our proposed approaches work for all these policies, but our discussions focus on the write-back policy due to limited space for our presentation.…”

“…As we describe the designs, we will also present experimental results as supporting evidence. We consider a set of block-level IO traces [AZ14] collected from a departmental private cloud as representative workloads. The characteristics of the traces are summarized in Table 3.1.…”

“…The flushing will also cause a surge in the storage server's IO load and affect the performance of the other VMs sharing the server. Second, the migrating VM needs to warm up the cache on the destination host, which may also take a long time, and it will experience substantial performance degradation till the cache is warmed up [HAWS13,AZ14].…”

“…It can be seamlessly deployed onto existing cloud systems as a drop-in solution and transparently provide caching and on-demand cache management. We evaluate it using a set of long-term traces collected from real-world cloud systems [AZ14]. The results show that RWSS-based cache allocation can substantially reduce cache usage and wear-out at the cost of only small performance loss in the worst case.…”