Feature

Disaster recovery relief

Ezine

This article can also be found in the Premium Editorial Download "Storage magazine: Storage managers give thumbs up to IP storage."

Download it now to read this article plus other related content.

TCP/IP WAN accelerators for replication

Requires Free Membership to View

TCP/IP WAN accelerators used for replication boost performance by three to 400 times the normal network speeds by shielding against bit error rates, jitter and TCP latency.

A number of vendors have enthusiastically attacked this problem and are now delivering cost-effective products. Each vendor's technology is innovative and varied. Technologies employed include packet-reordering, performance-enhancing proxies, compression, QoS and sophisticated flow control. The net results are impressive, with performance throughput increases between three and 400 times the normal network speeds. (See TCP/IP WAN accelerations for replication".)

The way TCP/IP accelerators work for DR and business continuity is simple, and requires nominal skills. First, an appliance is placed as a gateway or proxy on the LAN with the replication sender. Then another appliance is placed on the LAN at the location of the replication catcher. The appliances intercept and transmit all packets between the sender and catcher.

The products can be divided into two categories: those that work up to and above DS3 bandwidth (up to OC3, OC12, etc.) and those that only work up to DS3. Vendors in the "up to or greater than DS3" category include NetEx Software's HyperIP and Orbital Data's IP Express. Vendors in the DS3 and below category include Expand Network's IP Accelerator, ITWorx's NetCelera, Peribit's Sequence Reducers and Riverbed's Steelhead.

The one disadvantage to these solutions is the potential cost of putting appliances at a large number of remote sites. However, the savings in bandwidth costs and increased ability to hit DR and business continuity windows usually offsets the additional appliance and software license costs.

Virtual tape automation: Virtual tape has been around for years. Its success has been mostly in the mainframe market, where it's used as a nearline storage device that manages less-frequently needed data. Virtual tape appears to be stored entirely on tape cartridges, when some parts of it may actually be located in faster disk storage. By doing this, virtual tape makes tape more efficient and faster.

This never really caught on with the Linux, Unix and Windows markets because in those markets, tape is used primarily as backup and archival media. Automation makes virtual tape a low-cost methodology to increase the performance of DR and business continuity while lowering their TCO.

Automation works by making low-cost disk--usually parallel or Serial ATA--look like an automated tape library to the tape backup solution. This increases the backup speed to the speed of the disk array. From there, the backup data from one or multiple systems is written to tape automatically. Recoveries are faster because the data is organized more efficiently on the tapes and is easier to locate. (See "Automated virtual tape".)

Automated virtual tape

TCP/IP WAN accelerators used for replication boost performance by three to 400 times the normal network speeds by shielding against bit error rates, jitter and TCP latency.

This solution reduces the cost and complexity and increases the effectiveness of local DR and business continuity solutions. Products include ADIC's Pathlight VX: Integrated Disk-to-Tape Backup solution and Maxxan with FalconStor IPStor 4.0 Virtual Tape running within its MXV320 and SVT 100 intelligent switches. The ADIC solution includes the SATA array and the tape library. The Maxxan solution requires a disk array and tape library. These reduce costs and complexity, and increase the effectiveness of the DR and business continuity processes. But they don't address the TCP/IP bandwidth issues previously discussed when backup or replication must be carried out over long distances.

Leveraging these new solutions will require time, planning and effort. None of these products are a panacea. The applicability of any solution will vary by organization. Curing the pain may require the implementation of multiple solutions solving different pieces of the puzzle and working together.

How to proceed
Before evaluating, implementing or deploying any of these or other products, it's necessary to conduct an internal audit of your storage environment. Identify the sources of your disaster recovery pain. This may require the implementation of a storage resource management (SRM) tool. Once the pain is diagnosed, determine which product or products will eliminate or significantly reduce that pain. Design an evaluation pilot program to provide quantifiable benchmarks to measure against expectations. If the benchmarks prove to meet or exceed expectations, develop a sane deployment rollout program. Make sure there are measurable milestones and benchmarks along the way. Don't be afraid to require the selected vendors to provide SLAs. The SLAs should provide guarantees and financial penalties for failure to achieve those results. This process is not for the faint of heart; however, the relief of chronic pain is well worth the effort.

This was first published in April 2004

There are Comments. Add yours.

 
TIP: Want to include a code block in your comment? Use <pre> or <code> tags around the desired text. Ex: <code>insert code</code>

REGISTER or login:

Forgot Password?
By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy
Sort by: OldestNewest

Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to: