Copy management technology is quickly becoming a must-have for IT organizations that need to aggregate and manage their secondary data -- including backups, archive, and test and development environments -- and reduce storage costs.
But what factors should organizations consider before evaluating copy management systems?
Broadly speaking, there are two implementation models for copy management systems today. One is a hardware appliance with the software integrated. The other is software-only, where the customer acquires the hardware and installs the software.
As always, there are pros and cons to taking either path. Here are the main advantages of buying an appliance:
- Pretested, prevalidated hardware. The vendor has tested and validated that the hardware configuration provided works efficiently with the copy data storage software. This should ensure optimal performance, as well as iron out any compatibility bugs between the software, hardware components and firmware.
- Single support point. The vendor offers a single point of contact for both hardware and software support of the product. If any anomalies are found, the vendor can quickly test and (hopefully) resolve issues because it already has a test bed of the hardware in its lab.
- Predictable performance. The vendor has run tests and is able to provide data on how well its hardware performs. This can be important when designing and scaling out a copy management software product.
Of course, there are also disadvantages to purchasing hardware from the vendor. There may be a markup in cost compared to sourcing the hardware independently. For larger IT organizations, their own buying power may be as good as or better than the copy data storage vendor, so buying direct may offer better value for the money.
Buying direct also enables the IT organization to vary the hardware configuration, using newer technology (such as faster processors) before it is released by the copy data storage vendor. Using self-sourced hardware also enables the IT organization to use technology that has already been designed with operational standards for the organization. Introducing new hardware means writing new support processes.
As a rule, packaged products work well for organizations that do not have big hardware support ecosystems, as it is much easier to let the vendor be responsible for support and maintenance.
Copy management system features
So, what features should you consider across the range of products available? We highlight the seven most important features and how they can help influence picking the right copy management system.
Scalability. Typically, this refers to the amount of data that a product can store, plus the ability to scale out as required. There are several aspects to consider here.
First is the logical capacity the copy management product can accommodate, which may be limited in terms of storage, objects (virtual machines) or data (files). There also may be limits on the number of copies of each object retained (such as numbers of copies of specific data).
Physical capacity could be limited by the size of an appliance, plus the ability to scale out the product to multiple appliances or to increase storage capacity. The ability to support public cloud means the physical capacity of the product could effectively be unlimited, although the logical capacity could be a limiting factor.
Performance. Performance covers several aspects. At the ingest of data (either from a backup or from other sources), the ability to store data in the system is generally measured by throughput, such as megabytes per second or gigabytes per second. This important system limitation is reflected in scalability; can the copy data storage product scale to ingest larger amounts of data by adding nodes?
A second aspect to performance is the product's ability to support secondary access features, such as powering up virtual machines on the appliance itself. This kind of instant recovery is highly beneficial, but should not impact the continued ingestion of backup and other data.
Retention. Data retention describes how long individual backups or other data are retained by the platform. Ideally, data should be capable of indefinite retention (subject to policies) and should not be a limitation for design.
Retention policies determine how long data should be kept by the copy management system and must be complex enough to enable many different retention scenarios based on the data source and the time at which the data was ingested or identified.
If a copy management system simply manages the data copies on a storage platform, then, obviously, the limitation is on the underlying storage platform, and this may have an impact on product choice.
Integration. How does the product integrate with data sources? Most importantly, how does copy management integrate with existing hypervisor platforms and backup interfaces to take extracts of data? This kind of integration needs to ensure that data integrity is maintained (for instance, offering integration with the Microsoft Volume Shadow Copy Service for snapshot integrity).
Also consider integration from a storage platform perspective in areas such as managing data snapshots or where the copy management product does not actually extract data from the storage system.
Security. Imagine putting all of your key data assets in one place; security, at this point, becomes a real concern. First, consider security of data on the platform (whether or not it is an appliance). Data stored on the copy management system should be encrypted at rest and in flight, including to and from the public cloud (see cloud support below).
If copy data will be accessed through the backup system, then backed up data needs role-based access controls and audits. Roles-based restore access ensures that only owners of the primary data can delegate who can restore that data.
As copy management systems can expose the capability to restore data and virtual machines, this capability needs to be correctly secured with role-based access controls and, of course, audited.
Cloud support. If you have large volumes of data, writing everything to a local appliance is not cost-effective. IT organizations with large volumes of data can see significant benefits by offloading data to the public cloud, with caveats, of course.
Using the cloud as a data repository alone does not save much money, unless that data can be exploited from the cloud rather than through the appliance. This requires accessing metadata describing the data stored through a cloud-based mechanism, such as a virtual appliance.
Automation. Automation provides the capability to script and otherwise offload or integrate features of the copy management system. Much has been publicized recently about storage products that offer APIs, but automation could also be driven by command-line interfaces.
Automation enables scalability and integration with other orchestration systems. It also enables users to manage their data and recover or restore data without needing an administrator. This vastly simplifies the operational model for a copy management system, removing the burden of support from the backup or storage team.
Additional criteria for picking the right copy management system
Other considerations that have a bearing on choosing the right copy data storage product include the efficiency of the product. With large volumes of similar data, the ability to deduplicate becomes extremely important to managing cost.
Another consideration is data export and the degree of lock-in to using a copy management product. Although every platform has a degree of lock-in, you should consider the implications of moving your data to another platform.
Overall, once you understand your product requirements, you can map them to the features provided by the various copy management systems. Some requirements may be more critical than others.
The above list is in no particular order; however, prioritizing your features can help make product selection much easier. As always, a good set of well-defined requirements can help you compare the copy management systems on the market more easily.
How well do you know copy data management?
The benefits of having a copy management system in place
Backup trends focus on copy data management