Caching is the most important factor affecting the performance of a storage system. For write services, more caching means higher write merge rate and write hit ratio (the ratio of the same piece of data to be written again in the cache). For read services, more caches usually mean higher read hit ratios. At the same time, different types of services have different requirements for the cache. For sequential services, the cache is not large enough and only needs to meet I/O merging requirements. For random class businesses, a larger cache usually means better aggregation, which leads to performance gains.
SmartPartition
The core idea of SmartPartition is to ensure the performance of key applications by partitioning system core resources. Administrators can configure different cache partitions for different applications, and the system will ensure that the cache resources in this partition are exclusively owned by the application. It can also be combined with other QoS technologies, such as SmartQoS, to achieve better quality of service assurance.
Working principle:
The Cache partition technology isolates Cache resources required by different services to ensure service quality of certain critical services. SmartPartition can allocate cache partition resources of different sizes for different services (luns and file systems are actually controlled) to ensure service quality of key services.
Technical features:
- Intelligent partition control: SmartPartition automatically allocates cache resources based on the user-defined cache size and works with other QoS policies to maximize system service quality and ensure partition quality.
- Easy to use: SmartPartition is easy to configure and user-friendly. All operations take effect immediately. The system does not need to be restarted and partitions do not need to be adjusted.
