The term “solution” is not something to use lightly as it refers to a product or set of products that provide the total package. While no single vendor can be the answer to every pain and problem a business encounters, there are important items to consider before making an investment in a data protection solution.
This article will discuss the major items to consider before purchasing a data protection solution as well as the issues affecting businesses on multiple levels, including total cost of ownership and time spent on administration, maintenance, support and recovery.
More than just backup and recovery
When choosing a product to protect a company’s data, it must offer more than just backup and recovery. Sure, backup and recovery are important—cornerstones even—but they are not everything.
When talking about data protection, there is so much more to consider. Long-term data storage, otherwise known as data archiving, is not just another copy of a backup tape. All data protection solutions should have a way to separate data into classes that include, but are not limited to, backup data or archived data.
Backup data is usually short-term in nature (60 seconds to 60 weeks), while archive data is typically kept for six months to 60 (or more) years. Because of the difference in retention times, it is important to manage the media and the information differently. That’s not to say that the media should be set on a shelf with a label and forgotten about. On the contrary, long-term data must be actively managed.
When an archive is created and stored for several years, it is important to know the data can be retrieved when needed. Therefore, media needs to be checked on a periodic basis to verify the data can still be retrieved and used. To do this, a proper solution should be able to audit the media and ensure the data is viable.
Not only should it be able to verify the contents of the media, but it should be able to roll that data forward in time. In seven years, there will no longer be LTO5 libraries as it may be up to LTO10 or even some new technology, such as “data flux capacitor storage.” If there is a need to retrieve that old data and there is no longer an LTO5 drive, users may be in trouble. The solution needs to be able to roll data forward in time, moving the data from one type of media to a newer type of media, while keeping all metadata in tact as well as verifying the validity of the data. Without this feature, long-term data is stuck and married to the device that created it.
Archive data is only one example. There is also hierarchical storage management (HSM), application aware and encrypted data, and so much more. A true solution allows users to separate this data and treat different types of data in different ways. Without this ability, users are getting a backup and recovery product and not a data protection solution.
Subscription and support should not be forgotten
Tired of calling the data protection vendor and them saying the issue is a hardware problem only to have the hardware vendor point the finger back to the solution provider? Finger pointing is a huge waste of time and can actually cause major disruptions in service.
When choosing a data protection solution, look for one that offers peace of mind through world-class customer support and subscription and support (maintenance) contracts, which provide enormous financial value to customers.
For example, all improvements made to a solution in the past three years as well as the planned improvements for the next three years should be taken into consideration. This includes source and target data deduplication, an eight-fold increase in scalability and substantial improvements in reliability, performance and ease of use.
The best of both worlds occurs when all of these new features are provided to existing customers at no additional charge through subscription and support contracts. These customers enjoy all the benefits of new versions as they are released without incurring additional costs. As a result, subscription and support will be significantly enhanced during the lifespan of the solution.
Another factor that is just as important is the quality of support. If there are multiple products cobbled together to create a solution, it can cause a real problem as there is no single owner to any particular issue that occurs. The appliance model of data protection excels here as the provider can support the hardware, software, operating systems and everything in between.
When choosing a data protection solution, do not discount the value of support and subscription. Having a good support organization standing behind the solution can make life a lot easier.
Reliability: Key measure of a data protection solution’s ROI
At the end of the day, the real value of a data protection system lies in its ability to restore data when and where needed. If data fails to restore successfully, some part of the business is going to suffer, possibly with costly consequences. As a result, reliability is a key measure of a solution’s ROI.
A user may have the best, most expensive, fastest data protection solution in the world, but if data cannot be retrieved and used in a timely fashion, the point is moot.
When choosing a solution for data protection, look for experience in the data protection market—years spent protecting the business critical data of some of the largest organizations in the world. This is not a professional sports draft, so do not look for the diamond in the rough that can be molded into the perfect player (solution).
Instead, look for a solution that has a track record of being reliable over many years. The longstanding ability to migrate between storage devices as needed helps to ensure that organizations enjoy longevity on the platform, protecting their investments and saving on costs over the long-term.
Without this factor, all solutions become short-term fixes instead of long-term strategies that help businesses focus on future growth instead of today’s problems.
When choosing a solution, the ability to know the data can be recovered must be more important than any other factor. Yes, it is actually more important than price.
Users do not want to end up replacing the solution in just a few years (or less) because it cannot keep up with growth and continue to recover data.
Please, please make the data protection product easy to use
There are many data protection products on the market today, and all of them have features and functions that make them stand out. One of the major items to consider is how easy the solution is to use. When it comes down to “brass tax,” ease of use is one of the most important items. After all, the person responsible for data protection may not have a huge skill set or the time to spend on managing the solution on a day-to-day basis.
When choosing a data protection solution, look for the ability to manage the system from a single pane of glass. The user interface also needs to be simple enough that within a few minutes all daily tasks can be completed. As a bonus, find a solution that sends alerts to email and mobile devices in an easy-to-understand report that includes all important information about the previous night’s activity and the status of those activities.
This simple to use requirement dovetails into historical reporting. If the user knows what the solution is doing on a day-to-day basis, then he or she will also be able to tell what the system has been doing for the last few days, weeks, months or even longer. This allows for planning the future with little to no hands-on work. For example, if there is a report that gives weekly growth for the last 26 weeks, users will be able to predict when they are going to run out of space in the solution or when to purchase more tapes. It is a very simple example, but it shows that the solution should help plan for the future as well as operate today. And, following this example, users will be able to budget six months or more in the future for growth—a great advantage when budgets are tight.
The single-server footprint versus the master/media footprint can also make the solution easier or harder to manage as will automatic client software updates that keep IT administrators from spending valuable time making manual updates to systems across the infrastructure.
With business-wide administration, monitoring and reporting, plus the flexibility enabled by automation, a new solution should create administrative time-savings that can measurably reduce the cost of operations. If not, it is time to rethink the solution.
Remember this: A data protection solution should be the answer to the user’s problems, not cause more of them.
Data life-cycle management: From birth to death
IT organizations can drive up the cost of storage unnecessarily by treating all data the same and storing it all on the same media. All data is not created equal. When looking for a data protection solution, managing the ones and zeros from the time they are created to the time they can be deleted is very important.
Long-term archive and HSM allow organizations to store data on different tiers based on specific policies, enabling administrators to migrate and store data on the most appropriate tier. For example, older and less frequently accessed data can be moved to a slower, less-expensive storage platform, such as tape, leaving more expensive disk storage available for more high-value data. Automated data archiving also helps organizations ensure compliance with data retention policies and reduces the costs associated with compliance.
When making a decision about data protection, look only for solutions that help reduce costs by providing automated, policy-based data life-cycle management, moving data to the most cost-effective tier of storage while still meeting service level requirements. This will help ensure recovery objectives are met while enabling transparent data access. Do not be fooled by vendors who say that backup and recovery are enough.
Support for virtualized environments is key
Virtualization technology has helped IT organizations of all sizes reduce costs by improving server utilization and reducing application provisioning times. However, this produces two new problems that most people do not account for when choosing a data protection solution. First, the cost savings offered by virtualization can disappear quickly in the face of virtual machine sprawl. Second, the link between physical and logical devices becomes harder to map and track, making a virtual environment more complex than most can imagine.
Data protection can become a unique challenge in these environments. For example, backing up and restoring data for a dozen or more virtual machines residing on one physical server can bring all other operations on that server to a complete halt.
All that said, when searching for a data protection solution, be sure to investigate whether the product provides an effective solution to this challenge by eliminating the burden of running backups on a virtual machine and instead, off-loading backup workloads from a VMware ESX or ESXi-based server to a centralized vStorage backup server. The solution must improve the frequency of protection and enable faster recovery of data, helping increase the business value of virtualization. The solution should also help organizations reduce license management costs by removing agents from the individual virtual machines.
Data reduction, not just data deduplication
Data protection is not just about data deduplication. Many data protection products and providers talk about data deduplication as if it will save the world. In fact, data deduplication is only a small part of the solution. What needs to be talked about is across the board data reduction.
Data reduction technologies are the first line of defense against rapidly expanding data volumes and costs. A solution that provides built-in data reduction technologies, such as progressive-incremental backup, data deduplication and data compression, can enable organizations to reduce backup storage capacity by as much as 95 percent. Advanced tape management and efficient tape utilization capabilities can further reduce data storage capacity requirements.
While some solutions create massive amounts of duplicate data through repetitive full backups, necessitating expensive data deduplication solutions, others provide progressive-incremental backup technology that avoids the duplicate data in the first place by creating only an initial full backup and then capturing only new and changed data. Built-in data compression and data deduplication operate at multiple storage layers to minimize the amount of data being retained for operational and disaster recovery.
Reducing backup storage requirements dramatically not only helps reduce capital expenses, but can also decrease network bandwidth requirements and shrink backup windows. This results in reduced operational impact of backups and helps to ensure high levels of application uptime.
Stop the madness and look for more than data deduplication. Look for the entire package of data reduction technologies.
License correctly and save money
Can two of the same things have two different prices? Absolutely. Not only can they have different prices, but they can also be dramatically different.
In the last few years, data protection solution providers have started to offer something besides “core-” or “server-” based. When buying software consider all options. One of the newest options is capacity.
When purchasing a data protection solution, look for a company that offers pricing options that allow users to pay for solutions in the manner that makes the most financial sense. In the past, licensing models were based on the number and power of processor cores in the servers being protected. They also had cost advantages for organizations with relatively large amounts of data and a small number of servers, or for organizations with other software products licensed this way.
But, some also offer a capacity-based licensing option that allows organizations to pay for the software based on the amount of data being protected. This model has cost advantages for organizations that have a relatively large number of servers. It also helps eliminate licensing cost surprises when servers are added or cores are upgraded. The software includes tools that can help the organization make accurate budget forecasts.
The capacity-based model has particular value in infrastructures with multiple applications that require data protection solutions. Under the capacity-based model, these solutions are included at no additional cost. In addition, using advanced features, such as data deduplication, which reduces the amount of data being protected, can decrease the amount of data being measured against the license cost.
When looking for a backup and recovery or archive and retrieve solution, ask the vendor to show all of the different ways it can be licensed. Ask for a price for all options, and then make an informed decision.
Scalability–You grow, it grows
When keeping pace with growing data, a major concern for IT organizations in terms of storage and data protection is how the solution will handle the growth.
If a user’s business has grown its capacity by 40-60 percent in each of the past three years, and it now supports billions of data objects, they need a solution that grows with them. This growing of capacity may be outpacing the data protection solution and there is a need to find a way to scale the protection.
This growth can be handled, but the scaling must be done in a logical manner. There are three ways to do this:
Scale out: This usually means that new hardware and software will be added as further resources to handle the load from the growth. This involves a significant investment in new resources.
Scale up: This is where users add new software on existing servers, such as creating a second copy of an application on existing hardware. While this does cut the cost of new hardware, it assumes that the existing hardware can handle the load.
Scale in: If users can find a data protection solution that grows as they grow without additional resources, then they have hit pay dirt. There is usually no additional investment involved, however, users must have 20/20 foresight.
A solution should be on hardware that will grow into the future and have software with proven ability to grow at the same pace or better than the company’s growth.
When choosing a product for data protection, users should decide up front if they want to scale up, out or in. That up-front decision will dramatically change the amount spent down the road in years three and five.
Not all data is created equal. Stop treating it that way.
IT organizations can drive up the cost of storage unnecessarily by treating all data the same and storing it all on the same media. Let’s face the fact: a resume is not as important as the payroll database or even the email database. So, why do IT folks use the same storage policy for both?
Stop using one policy to rule all data. It might be simple, but it will kill the bottom-line. When looking for a data protection solution, find one that allows policies to be set to treat data differently.
Important data should be prioritized as tier one data that gets backed up most often and most quickly. Perhaps that data can stay on disk for fast restore.
Everything else that is not business critical is considered tier two or junk data. Tier two data should go directly to tape and be stored since it is not business critical. Junk data, like photos or temp files, can be deleted.
Tier two data is a great target for HSM, which allows organizations to store data on different tiers based on specific policies, enabling administrators to migrate and store data on the most appropriate tier. For example, older and less-frequently accessed data can be moved to a slower, less-expensive storage platform, such as tape, leaving more expensive disk storage available for more high-value data.
When considering the various data protection solutions on the market, look for one that helps reduce costs by providing automated, policy-based data life-cycle management, moving data to the most cost-effective tier of storage while still meeting service level requirements. This helps to ensure recovery objectives are met and transparent data access is achieved. Automated data archiving also helps organizations ensure compliance with data retention policies and reduces the associated costs.
Find a data protection that allows data management while continuing to operate in a simple, easy to understand environment.
Think about it now or pay for it later.
Recovery: A unified approach
Is your organization using different products to protect different types of data or different systems? If so, start thinking about standardizing on a single product. Think of all the time, training and resources that will be saved.
Unified recovery management (URM) brings under one user interface the ability to manage data protection throughout the business, supporting different applications and types of data on different operating systems in different locations and with different policies and backup requirements. From a single point, administrators can manage multiple data protection and recovery tools, including diverse solutions that are dedicated to different tasks. It helps eliminate the costs and complexities associated with deploying and managing multiple point solutions.
When looking for a company that provides data protection, look for one that simplifies and streamlines storage management, helping organizations control both the risks and costs of data protection and recovery. With fewer “moving parts” for managing the various solutions in operation, administrators can ensure faster, more reliable backup and recovery processes. The solution also provides built-in replication for highly available disaster recovery, helping to reduce downtime and the business costs that can result. These process improvements contribute to higher levels of service, making it easier for organizations to meet service level agreements.
The ability for a single person with very limited knowledge to manage an entire businesses’ data protection solution is important. With a unified approach, users gain the ability for IT staff to be nimble and forward-thinking. No longer are they in “reactive mode,” but they start to exist in “proactive mode.”
This is important because it saves time and money.
Jarrett Potts, director of strategic marketing for STORServer, a leading provider of proven data backup solutions for the mid-market. Before joining the STORServer team, Potts spent the past 15 years working in various capacities for IBM, including Tivoli Storage Manager marketing and technical sales. He has been the evangelist for the TSM family of products since 2000. www.storserver.com