Application Scenarios
This section describes application scenarios of SmartPartition. SmartPartition allocates cache resources on demand for mission-critical applications and high-level users.
Ensuring the Performance of Core Services in a Multi-Service System
As the performance and capacity of a storage system keep growing, multiple applications are typically deployed in the same storage system to simplify the storage architecture and reduce configuration and management costs. However, those applications contend for storage resources, seriously affecting the performance of each service. Based on service characteristics, SmartPartition can allocate different cache partitions to different services, so that cache resources in a SmartPartition are exclusive to the corresponding service, thereby meeting the performance needs of different services and ensuring that mission-critical services run smoothly.
For example, a production system and its test system are running on the same storage system. Table 1-2 describes the service characteristics of the two systems.
Service Type |
Characteristics |
---|---|
Production system |
Frequent data reads and writes |
Test system |
Frequent data reads and rare data writes |
By using SmartPartition, you can allocate cache partitions to the production system and test system, enabling the two systems to own cache space independently. Then, appropriate read and write cache sizes are set based on the two systems' I/O read and write frequencies, improving the access performance of the production system without affecting proper running of the test system. For example:
- SmartPartition policy A: SmartPartition 1 is created for the production system. (For example, the read cache size is 2 GB and the write cache size is 1 GB.) The read and write cache sizes are sufficient for the production system's frequent I/O reads and writes.
- SmartPartition policy B: SmartPartition 2 is created for the test system. (For example, the read cache size is 1 GB and the write cache size is 512 MB.) The read and write cache sizes set for the test system are limited, ensuring proper running of the test system without compromising the performance of the production system.
Meeting Service Quality Requirements of High-Level Users in a VDI Scenario
In a virtual desktop infrastructure (VDI) scenario, different users have different application QoS requirements. The data center is faced with the dilemma of meeting each user's requirements while bringing existing resources into full play.
SmartPartition allows you to allocate different cache capacities to different users. When resources are limited, priority is given to higher-level users, ensuring their QoS.
For example, the storage resources of a data center are provided for multiple users. Table 1-3 describes the QoS requirements of users A and B.
User |
QoS Requirement |
---|---|
User A (gold level) |
High |
User B (silver level) |
Low |
Based on SmartPartition, the storage resource provider can create one cache partition for user A and one for user B and configure different read/write policies for them.
- SmartPartition policy A: SmartPartition 1 is created for user A. (For example, the read cache size is 2 GB and the write cache size is 1 GB.) The read and write cache sizes ensure that user A's applications provide outstanding read/write performance.
- SmartPartition policy B: SmartPartition 2 is created for user B. (For example, the read cache size is 1 GB and the write cache size is 512 MB.) The read and write cache sizes for user B is limited, ensuring proper running of user B's applications as well as high QoS of user A's applications.