InfoScale™ 9.0 SmartIO for Solid-State Drives Solutions Guide - Linux

About SmartIO FEL-based writeback caching for applications running on VxFS file systems- DLV 14 and later

InfoScale Storage and InfoScale Enterprise supports write-back caching on solid-state drives (SSDs) for applications running on Veritas File System (VxFS) file systems. In case of DLV 14 and later, application writes are cached in a Front-End Log (FEL)-based writeback device.

The FEL-based writeback caching mechanism is functionally similar to that of the SmartIO write-back cache. However, many aspects of the FEL-based design is expected to result in better performance than SmartIO write-back caching. Also, the SmartIO FEL-based write-back caching removes the restriction of two node cluster, which was present in the SmartIO write-back cache. Since both the caching types are functionally similar, the FEL based write-back caching is a replacement for the SmartIO write-back caching, on DLV 14 and later. FEL-based write-back cache continues to use many of the user management interfaces as the SmartIO write-back cache. For older disk layouts i.e. DLV 13 and prior, SmartIO write-back caching mechanism remains unchanged.

An FEL-based write-back cache is a volume that is separate from the primary volumes of a filesystem which is used as a persistent staging area, allowing synchronous writes to return quickly to the user, regardless of the complexity of the write itself. Writes to VxFS can potentially require a lot of processing, depending on the features in use, before they are written persistently to disk. For example, write to a sparse region of a file requires allocation in the context of the write itself, which can introduce significant delay to the synchronous write. It is difficult for a user or application to predict this delay ahead of time. The FEL-based write-back caching helps to buffer this variability, providing minimal and predictable write latency.

In writeback mode, an application write returns success after the data is written to the cache, which is usually on an SSD. At a later time, data is flushed to the primary storage. Write order fidelity is not guaranteed while flushing the dirty data to the disk.

The writeback caching mode gives good performance for writes, but also means that the disk copy may not always be up to date. Because data from successful writes might only exist persistently in FEL, until it is flushed, availability of the FEL volume is as important as that of the main filesystem volumes. If a cache device fails, a file that is cached in writeback mode may not be completely present on the disk. SmartIO has a mechanism to flush the data from the cache device when the device comes back online.

Storage Foundation Cluster File System High Availability (SFCFSHA) provides additional protection from data loss by mirroring the FEL cache between nodes using FSS. The filesystem can recover from a node failure by reading the FEL from a mirror on another node, resulting in full access to the filesystem without requiring the failed node to come back.

In case of SFCFSHA, when writeback caching mode is enabled, SmartIO mirrors the write-back data at the file system-level, to the other node's SSD cache. This behavior, called cache reflection, prevents loss of write-back data if a node fails. In case of a node failure, the other node flushes the mirrored dirty data of the lost node as part of reconfiguration. Cache reflection ensures that the write-back data is not lost even if a node fails with pending dirty data.

In the case of local mount, if there is no cache mirror, the disk copy of the file can be incomplete or stale if the FEL cache fails.

After the FEL-based write-back caching is enabled on the mount point, the qualified synchronous writes in that file system are cached. SmartIO determines if a write qualifies for write-back caching, using criteria such as the following: