Solid state drives (SSDs) can be used in a variety of way within OneFS. Prior to SmartFlash, or L3 cache, SSDs were used exclusively as file system devices. SmartPools provided the mechanism to use these SSDs primarily as metadata acceleration devices, but also for reducing latency of actual data reads and writes for the appropriate workloads.


The introduction of SmartFlash provided a large, cost-effective method of extending of main memory per node from gigabytes to terabytes. This allowed clients to retain a larger working set of data in cache, before being forced to retrieve data from spinning disk via a higher latency HDD operation. The L3 cache is populated with “interesting” L2 blocks that are being dropped from memory as a result of L2’s least recently used cache eviction algorithm. Unlike RAM based caches, since L3 is based on persistent flash storage, one the cache is populated (warmed), it’s highly durable and persists across node reboots, etc.


The table below provides a comparison of L3 cache with the other OneFS SmartPools SSD usage strategies. The principle benefits of L3 cache are around metadata read activity, user data read activity, and assistance with job engine performance. While L3 cache does not directly improve write performance, offloading random reads to SSD does have the additional benefit of freeing up the hard drives for write I/O. L3 cache is implemented at the node pool level and is designed to be simple to use.

 

In contrast to L3 cache, with SmartPools data on SSD strategy only the files specifically targeted to SSD benefit from the increased read and write performance. The remainder of the data on the node pool lives exclusively on hard disk and will not benefit from SSD. SmartPools Data on SSD strategy is licensed functionality and is typically configured using file pool policies, or, occasionally, manually via the ‘isi set’ command.

 

Compared to L3 cache, SmartPools SSD strategies in general typically require more complex configuration and must be monitored closely so as not to exceed the available SSD capacity.

 

The table below provides a comparison of L3 cache with the other OneFS SmartPools SSD usage strategies. The principle benefits of L3 cache are around metadata read activity, user data read activity, and assistance with job engine performance. While L3 cache does not directly improve write performance, offloading random reads to SSD does have the additional benefit of freeing up the hard drives for write I/O. L3 cache is implemented at the node pool level and is designed to be simple to use.

 

In contrast to L3 cache, with SmartPools data on SSD strategy only the files specifically targeted to SSD benefit from the increased read and write performance. The remainder of the data on the node pool lives exclusively on hard disk and will not benefit from SSD. SmartPools Data on SSD strategy is licensed functionality and is typically configured using file pool policies, or, occasionally, manually via the ‘isi set’ command.

 

Compared to L3 cache, SmartPools SSD strategies in general typically require more complex configuration and must be monitored closely so as not to exceed the available SSD capacity.

 

SSD_strategy_comparison_1.jpg

              

     * Data located on SSD benefits. Other data on node pool does not benefit.