It can be difficult to gain perspective on storage growth and how to manage it. One of the most common mistakes end users make is deciding to purchase storage at the end of a lease contract or when there is a large development project looming on the horizon. The added pressure can result in short cuts in product evaluation and can also lower the end user’s negotiating power with vendors. So what can end users do to avoid these pitfalls?
Planning ahead is the first step, says John Lallier, vice president of technology at FalconStor, but there are things end users can do even if they find themselves in a jam.
“If an end user is in this situation, they should not compound the problem by committing to purchase storage and storage services from the same disk vendor,” says Lallier. “When using storage services from a hardware vendor, these services work only with the disks in the vendor’s storage array, resulting in vendor lock-in, leaving the user with no room for negotiation.”
Lallier says end users should purchase storage services from a neutral party — that way, they can negotiate with a large pool of disk vendors, purchasing only capacity and avoiding vendor lock-in.
Eran Farajun, executive vice president of Asigra Inc., says that as soon as the original purchase is made, end users should specify that the storage infrastructure be open so that data can be moved to an alternative platform. “End-users can also purchase storage infrastructure from more than one vendor, and introduce a software layer above the storage to manage it,” he says.
End users also tend to become locked into a specific vendor because they purchase a homogeneous solution that was “sold” to them as a heterogeneous solution.
Zophar Sante vice president of marketing and development at Sanrad, says that there are few truly “open” solutions, so users continually return to the same manufacturer for additional services and products. He believes it is very important for end users to understand what is available, what is reliable, and what will work within their environment.
Another suggestion Sante offers is that end users stay abreast of what’s new in the storage arena by continually researching new companies and technologies. “The key for end users is to call the vendor directly and discuss their specific requirements with the vendor’s solution architect or engineer,” he says. “Most storage vendors have technical people who can assist in determining if a specific application will run on a specific platform.”
Sante also suggests that end users get references from the vendor of other end users who purchased that same product to make sure that the product works as advertised. And, he says, end users should always request a money-back guarantee from the vendor so that if the product does not perform as advertised, it can be returned without a hassle. In other words, get the product in-house and try it. Sante suggests that the best negotiating stance end users have is to get a competitive product or technology in-house and see if it will do the job.
Performing A Business Impact Analysis Is Also Key
Performing a business impact analysis (BIA) before purchasing a storage solution can help end users improve business continuity. Lallier says a BIA is definitely a good idea, and adds that end users should have three main goals in mind when performing a BIA:
- Make sure the storage solution is compatible with the existing storage infrastructure (HBAs, switches, etc.).
- Qualify the storage solution to ensure that it is reliable and matches the SLA of the intended applications.
- Make sure that the performance of the storage solution is adequate for the applications.
Scott McLeod, director of professional services for MaXXan, agrees that performing a BIA is the way to go, and says that end users should identify and quantify risks, prioritize mitigating factors and classify associated data and devise appropriate counter-measures.
Page 2: Backup and Recovery is Another Planning Issue
Backup and Recovery: Another Planning Issue
Designing backup and recovery architecture is another important storage planning issue.
Peter Hunter, product marketing manager at EqualLogic, says the first step in building a backup and recovery architecture is to implement an infrastructure capable of isolating and minimizing data movements. The foundation to this strategy starts with a SAN.
Data, Hunter says, can move more freely in a networked storage environment, and, by using snapshots, application servers can be liberated from the task of having to stream large I/O operations from the application server to the backup device or backup server. Instead, the application server simply requests that the storage array take a snapshot — an instant point-in-time copy of the data that resides on the SAN itself — and then the backup server mounts the snapshot and moves it to a tape device. The application server is never burdened with having to move the actual data, greatly improving backup operations, and administrators gain the option of snapshot recovery from disk, the fastest of all backup recovery techniques.
Backup has traditionally been a labor-intensive and time-consuming process, with a lot of data moving from servers to tape libraries. With snapshots there is less data movement, turning backup servers into automated data coordinators. “Snapshots, by their nature as a point-in-time image of the database that captures only the changes to the data at any given moment, are lightweight, easy and inexpensive to store, even over long distances, and restoration from snapshots is instantaneous, reliable, and accurate,” says Hunter.
Snapshots can be replicated offsite to geographically isolated locations for maximum data protection. Disaster tolerance through replication is now accessible to a broader market with the adoption of iSCSI, the storage protocol based on commodity IP technology, Hunter says. “No longer are expensive SAN bridging equipment and specialized storage networking expertise required,” he says.
Robert MacIntyre, vice president of business development and marketing at Netex, believes that storage vendors are on the right track in proving best of breed backup and recovery applications capable of operating over IP infrastructures. “IP is the network of today as well as the network of the future,” he says.
MacIntyre suggests that storage vendors focus their development resources on providing the right application features for their customers. He believes doing so will eventually make storage more affordable and accessible. “Many storage vendors have attempted to solve transport issues as part of the storage offering,” he says. “I believe that storage vendors should concentrate on storage-related features and leave the transport issues to vendors that are best suited for enhancing application performance and data movement between locations.”
In order to predict future storage requirements and let management accurately plan for future storage purchases, it is vital for end users to be able to calculate a server’s storage resources. There are other concerns as well, including which storage functions should an end user outsource, and which should remain in house. What are the best practices and strategies for designing, implementing, and managing storage of digital archives? How is the storage industry changing and what are the important trends? The final part of this series will answer those questions and also offer views on best practices for managing storage infrastructure.