IBM XRC Performance Monitor
Providing the data to meet XRC desired service levels
Extended Remote Copy (XRC), also known as z/OS Global Mirror is IBM's premier disaster recovery option for zSeries storage systems.
The XRC Performance Monitor provides additional information to better manage an XRC environment and help XRC achieve the desired performance (typically measured as the latency in updating the mirrored system)at a low cost. The monitor provides information in realtime and also maintains historic records. From a very high level or from a detailed perspective, administrators can view the parameters that contribute to system performance.
Several parts of an IBM zSeries™ environment affect the performance of XRC. The XRC Performance Monitor gathers information from important sources. This functionality not only helps determine the optimum XRC configuration,but also helps quickly identify performance bottlenecks. The XRC Performance Monitor is also useful in spotting trends that may lead to future performance problems;thus it can also assist with capacity planning.
- Provides essential information to verify that Extended Remote Copy (XRC)is meeting the desired service levels
- Collects the data needed to help optimize the use of resources and minimize costs
- Aids in capacity planning with historical trend information
- Helps eliminate performance bottlenecks quickly with real-time information
- Monitors key parameters on an ongoing basis and provides exception reports
- Extended Remote Copy
The IBM Extended Remote Copy (XRC) facility is a disaster recovery solution that helps provide fast recovery of computer operations if a site fails. It provides copy services to help maintain a secondary site with mirrored data so that a switch over can take place in a short amount of time. The primary site is connected to a secondary site using channel extenders and a telecommunications link (often multiple T1 or T3 lines).The System Data Mover software (frequently running on the secondary host) controls the transfer of data from the IBM TotalStorage Enterprise Storage Server™ (ESS) at the primary site to the disk subsystem at the secondary location.
- The XRC Performance Monitor is accessed using any terminal connected to the host running the System Data Mover software. This can be either the primary or secondary host
- Components of the XRC Performance Monitor
The XRC Performance Monitor comprises three separate modules, which are integrated under one user interface.
- History monitor summarizes how the XRC system was running at a prior point in time. At approximately 15-minute intervals, this report displays key performance information in an easy-to-read sorted table. Detailed information can be obtained about each reader and specific volumes. This information, displayed graphically, can help identify volume contention that may have created bottlenecks. This report includes details about the telecommunications link, such as the amount of data transferred, as well as measures of the mirrored site, such as peak delay times and average delay times.
- Realtime monitor provides a summary of how single or coupled system data movers are running (system data movers update the secondary storage devices).This report queries the Interactive System Productivity Facility (ISPF)to obtain the results. By displaying realtime information, the XRC Performance Monitor enables administrators to detect and resolve problems quickly.
- Batch exception monitor checks the monitor information at user defined intervals for predefined thresholds. This report generates console messages whenever thresholds are exceeded so administrators can take action, therefore eliminating the need for someone to constantly check the monitor.
Optimize the telecommunication link
The bandwidth of the telecommunications link between the primary and secondary sites is one of the key manageable resources in an XRC system. Because of the cost of high-speed circuits, overbuying capacity can add expense without much benefit. Likewise, if the link is too slow, unacceptably long delays may be introduced when updating the mirrored system.
One of the key applications of the XRC Performance Monitor is to make sure the link is sized properly. Typically, traffic on the telecommunications link varies throughout the day. For example, batch updates to an online system at night often produce much higher traffic on the XRC link than daytime operations. Unattended charting enables administrators to see latency as it varies throughout the day so that they can size resources accordingly. Historical charts show trends, which can be useful for predicting future demands.
For more information, contact your IBM representative or IBM Business Partner