Excel spreadsheets were not intended for storage administrators tracking capacity utilization and needs, but targeted rather towards accountants and others working with numbers. Out of necessity, however, spreadsheets have become the tool of the trade for understanding the storage landscape. Many EMC customers tell me, even with this method, they are still flying blindly with only a stove pipe view of their storage use and availability.
Data growth and cloud computing are going to change this behavior, though. The latest IDC Digital Universe Study estimates that data will double every two years between now and 2020. This growth underscores the importance of getting the most out of storage investments. With about 93% of all organizations using or planning to use server virtualization, storage is the next big opportunity for automation. Software-defined storage offers the promise of storage abstraction, pooling, and centralized management but in the meantime, some might think makeshift solutions like using spreadsheets have to suffice. That is not true. There are solutions available today for getting a data center wide view of storage use, availability, and performance.
The Data Challenge
The Enterprise Strategy Group’s (ESG) Top Ten Most Important IT Priorities (2012) listed the top three (3) initiatives to be: 1. Increase use of server virtualization, 2. Major application deployments of upgrades, and 3. Manage data growth. Taking these priorities one-by-one, it is no surprise that server virtualization continues to expand with the adoption of cloud models. Similarly, with a virtualized compute foundation, moving more applications into the virtual environment makes sense—as does ensuring their performance. Lastly, data growth—or more specifically, how to deal with it—is important to getting the most out of applications in the virtual environment.
Data growth means large amounts of data stored in SAN and NAS storage, and increasingly object storage. Without a full view though, existing storage can get underutilized—or enough capacity may not be available when needed. As a result, maintaining existing services or delivery new services at the appropriate cost point becomes increasingly difficult. For those data centers still following the spreadsheet method, the process can be tedious and the results prone to human error. It is also inconsistent that an automated application and compute environment depends on a mostly manual process for storage management.
Vendor-specific element management solutions only provide a view into a specific array type, meaning multiple monitoring and management solutions are often needed for each vendors’ arrays. Additionally, with dynamic virtual machines (VM) and applications, these storage virtualization and management solutions do not provide the full view from the application and server/VM layer to LUNs or file shares.
These disparate monitoring and management tools can negatively impact the applications that manipulate and present data in a cloud environment. Common issues include:
- Lack of end-to end visibility into application dependencies
- Difficulty in understanding storage usage trends to manage capacity effectively
- Inability to ensure configuration compliance
- Limited overall system-wide performance information
Getting Into Focus
Recently, I was talking to a data center manager about his storage systems in his private cloud model. Though attentive to the discussion about storage automation, what piqued his interest was the information he could extract from existing storage. He was intrigued by the possibility that he could get an end-to-end view of his environment, including storage, with less tools and make real-time decisions based on a more complete picture of his data center.
Think about it. If this data center manager or you could compile an aggregate view of your multiple block and file storage systems, you could, in effect, pool the capacity for management purposes already today. This means that, for example, your EMC VMAX block storage with a maximum 4 petabytes (PB) capacity per array could be viewed with other EMC VMAXs for a maximum capacity many times higher. Additionally, with a system-wide view, applications could have a better chance of getting the best storage type for the processes they support.
Introduced this past summer, EMC Storage Resource Management Suite is a good example of a cross-domain software management tool set for service assurance in the cloud. It includes EMC ProSphere (now release 1.7) for storage management, EMC Storage Configuration Advisor for configuration management, and EMC Watch4net for performance monitoring and reporting.
The Storage Resource Management Suite (Suite) provides a comprehensive end-to-end analysis and reporting of heterogeneous environments that include EMC and third-party (e.g. IBM DS, IBM SVC, NetApp, and HP) block, file, and virtualized storage. The Suite provides a better understanding about storage capacity utilization and array performance. Users also gain insights into storage service levels to better leverage inherent array capabilities such as EMC FAST VP. Organizations leverage topology and relationship mapping to analyze storage data paths from hosts or VMs to LUNs and file shares. Key capabilities include:
- Application-level visibility: While oriented to storage administrators, the Suite provides views into compute and networking to help data center administrators in many areas to ascertain how changes impact other areas. Report packs like for Oracle, for example, give storage administrators a view into database availability and performance.
- Optimized storage resources: In addition to usage and availability for EMC and third-party storage, the Suite provides visibility of array-specific capabilities like EMC FAST VP and EMC Federated Tiered Storage (FTS) data to ensure application performance service-level agreements (SLAs) are met.
- Accurate configuration management: The Suite validates any configuration change against the EMC Support Matrix or internal best-practices as well as the ability to monitor and alert on any violations before they impact applications.
- Performance-monitoring: System-wide trending and alerts based on key storage metrics ensure no SLAs are out of compliance.
Given this breadth of storage and configuration management and reporting, it’s a wonder why anyone would want to do this process manually. For our data center manager, the EMC Storage Resource Management Suite means he needs fewer tools with support for heterogeneous arrays, though he gets more uses because the Suite spans domains, offering useful insight into the storage data path and database availability and performance.
The portfolio keeps expanding too. Most recently, the Storage Resource Management Suite added support for EMC Isilon and EMC VPLEX systems. These additions mean more possible file-based NAS storage systems and support for active-active federated environments—all part of the cloud taxonomy. Given this breadth of EMC and third-party array types, might object storage be next? Stay tuned as EMC continues to automate storage in the data center, making Excel spreadsheets for storage management obsolete.