By using Server Priority Manager, you can designate prioritized ports (for example, for production servers) and non-prioritized ports (for example, for development servers) and set upper limits and thresholds for the I/O activity on these ports to prioritize I/O operations to host servers requiring high-throughput I/O operations and prevent low-priority activities from negatively impacting high-priority activities.
If you want to use Server Priority Manager to control the I/O priority for iSCSI-attached volumes, you must use CCI to issue SPM commands to the volumes (LDEVs). Also, if you want to use SPM to control the I/O priority for individual FC-attached volumes, you must use CCI to issue SPM commands to the volumes. When you use CCI to issue SPM commands to volumes, you can specify an LDEV and a WWN or iSCSI name of a host adapter.
- Specifying ports (GUI or CLI)
- Specifying LDEVs (CLI only)
Performance of high-priority hosts
In a SAN environment, the storage system is usually connected with many host servers. Some types of host servers often require higher performance than others. For example, production servers such as database and application servers that are used to perform daily tasks of business organizations usually require high performance. If production servers experience decreased performance, productivity in business activities can be negatively impacted. To prevent this from happening, the system administrator needs to maintain the performance of production servers at a relatively high level.
Computer systems in business organizations often include development servers, which are used for developing, testing, and debugging business applications, as well as production servers. If development servers experience decreased performance, development activities can be negatively impacted, but a drop in development server performance does not have as much negative impact to the entire organization as a drop in production server performance. In this case, you can use Server Priority Manager to give higher priority to I/O activity from production servers than I/O activity from development servers to manage and control the impact of development activities.
Using Server Priority Manager you can limit the number of I/O requests from servers to the storage system as well as the amount of data that can be transferred between the servers and the storage system to maintain production server performance at the required levels. This practice of limiting the performance of low-priority host servers is called upper-limit control. The upper-limit control is automatically disabled when traffic between the servers and the storage system drops to user-defined levels called thresholds.
While upper-limit control can help production servers to perform at higher levels during periods of heavy use, it may not be desirable when production servers are not busy. For example, if the I/O activity on production servers is high between 09:00 and 15:00 hours and decreases significantly after 15:00, upper-limit control for development servers may not be required after 15:00.
To address this situation Server Priority Manager provides threshold control in which upper-limit control is automatically disabled when I/O traffic between production servers and the storage system decreases to a user-specified level. This user-specified level at which upper-limit control is disabled is called the threshold. You can specify the threshold as an I/O rate (number of I/Os per second) and a data transfer rate (amount of data transferred per second).
For example, if you set a threshold of 500 I/Os per second to the storage system, the upper-limit controls for development servers are disabled when the I/O rate of the production servers drops below 500 I/Os per second. If the I/O rate of the production servers goes up and exceeds 500 I/Os per second, upper-limit control is restored on the development servers.
If you also set a threshold of 20 MB per second to the storage system, the upper-limit controls for development servers are disabled when the amount of data transferred between the storage system and the production servers is less than 20 MB per second.
Requirements for using performance functions
- License keys for performance management: The license keys for the following software products must be installed on the storage system:
- Performance Monitor
- Server Priority Manager
- Virtual Partition Manager
For details about installing license keys, see the System Administrator Guide.
- Access privileges for
Device Manager - Storage Navigator: Administrator access for
Device Manager - Storage Navigator or write access for the performance management software products is required to perform operations. Users without Administrator access or write access can only view the performance management information and settings. You need specific administrator roles to use the following functions:
- Performance Monitor: Storage Administrator (Performance Management)
- Server Priority Manager, Virtual Partition Manager: Storage Administrator (System Resource Management)
- Java: Java is required to use Server Priority Manager on the Device Manager - Storage Navigator computer. For details about installing Java and configuring Device Manager - Storage Navigator, see the System Administrator Guide.
- Secondary windows on
Device Manager - Storage Navigator:
You must enable secondary windows if you plan to use any of the following functions in Device Manager - Storage Navigator (HDvM - SN):
- Login Message function
- Data Retention Utility
- Server Priority Manager
- Compatible PAV
- Compatible XRC
- Volume Retention Manager
- Cache memory for Virtual Partition Manager: Use of Virtual Partition Manager might require additional cache memory in your storage system.