ViPR SRM Data Normalization

One of the greatest features of ViPR SRM 3.5 is the ability to look at your entire storage environment as a whole. This manifests in the ability to understand end-to-end relationships (for example, in topology views, path details and chargeback reports) and to see global views that aggregate all of your storage. Prior to ViPR SRM 3.5, this functionality was only available for EMC VMAX and VNX arrays. An important new feature set in v3.5 is that these capabilities are being extended to other top storage platforms, from other vendors in addition to EMC.

In order to accomplish this, the ViPR SRM engineering team goes through an analytic process called “data normalization.” To create end-to-end relationships within ViPR SRM, we need to correlate data from all the devices in the path, such as hosts, hypervisors, switches and arrays. However, each of the APIs for the various devices has a different way of representing the data used for this correlation. So, for a new array type, we have to modify this data to make it consistent with the representation by other devices. Similarly, for global dashboards that aggregate capacity across all storage, derived metrics need to be created for each new array that meet a consistent definition for each capacity classification, so that only like values are added.

Normalization starts with a thorough engineering analysis for each array type, and then the dedicated expert teams for topology, global reports, data modeling and each array platform work together to create the integrated solution.

In v3.5, we are bringing new normalized capabilities to additional EMC platforms including EMC VPLEX, Isilon and XtremIO,as well as platforms from other vendors such as NetApp, HDS, HP StorageWorks P9000, and IBM XIV, with other platforms in the pipeline for future releases.

Below is a simple topology map that shows the relationship dependencies from the host to the storage, which provides an example of the type of end-to-end capabilities we are adding for new arrays.
gary_image_1The map can be expanded to see more detail as you select the different components along the data path. As you make additional component selections, you are presented with different reports such as performance from the host, switch or storage perspective. You can view the path details, storage connectivity, and capacity. This level of visibility helps you troubleshoot performance problems. So rather than going into multiple element managers to piece together a report that either proves whether you do or don’t have a problem, this can all be done within ViPR SRM. And if you need to create a report to prove the storage infrastructure isn’t the culprit of the problem, creating a basic custom report is as simple as selecting the reports you need, saving it and emailing it.

Here’s another expanded topology view, this time for an IBM XIV array.
gary_image_2
Additionally, we are adding new arrays into global reports and dashboards. For example, the capacity of these arrays is being added to the Enterprise Capacity Dashboard.

gary_image_3
You can also view your storage all at once in other global views, such as Explore, where you can see your entire storage inventory.

gary_image_4
In Usable Capacity by Pool, you can see the capacity of storage pools across all of the supported array types from one view.

gary_image_5

The Array by Service Level report below provides a view into service level capacity in terms of used, free and the used trend.
gary_image_6
These are just a few of the examples of the enhanced views that are normalized across many storage types in ViPR SRM v3.5.

Should you wish to learn more about ViPR SRM 3.5 features, we invite you to visit the community.

About the Author: Gary Roberts