Kashya has unveiled a data protection appliance that costs just a fraction of competing systems from EMC, Hitachi, and other top-tier vendors.
As if that weren't enough, the start-up also claims its KBX4000 appliance offers greater performance and features without locking the user into proprietary technology.
Pricing starts at about $25,000 per site, while Kashya reports a competitive system from EMC would cost about $200,000 for a 0.5 terabyte system — SRDF/TimeFinder software for about $65,000, and another $120,000 - $150,000 for the necessary hardware. Kashya says its appliance doesn't require additional hardware until about 10-12 terabytes, at which point the KBX4000 costs about $185,000, compared to $445,000 for an EMC system.
Among the features Kashya packs into the KBX4000 are policy-based controls, bandwidth optimization, and the ability to replicate data for any distance in a heterogeneous storage environment, thus enabling customers to preserve their current infrastructure investment.
"For years, the customer has been looking for a way to non-disruptively transfer data over long distances without impacting application performance and without being concerned with data consistency," says Arun Taneja, founder of and consulting analyst for The Taneja Group. "Kashya is not only delivering on this promise, but has added additional features such as multiple replication policies and the ability to assign value to data via simple policy statements. This is very impressive indeed."
Tony Asaro of the Enterprise Storage Group's ESG Lab notes a number of potential positives in Kashya's technology. "If what they say is true, Kashya has developed enterprise-class remote mirroring, which is complex technology," Asaro says.
Kashya supports asynchronous remote mirroring, enabling long-distance copies with reliable recovery capabilities, Asaro says. They support heterogeneous storage, so the secondary site can use less expensive storage and customers can use just one tool to do remote mirroring for all of their storage.
The technology is out of band, which means it is not in the data path and storage vendors can't say they don't support it, Asaro adds. It uses the IP network, which can reduce costs by minimizing infrastructure costs and using less expensive WAN links, and it compresses data, which means customers can reduce bandwidth requirements on the WAN links.
A monitoring tool determines the bandwidth requirements of the customer, so customers can figure out what size WAN link they need, and they can also set policies for different types of data to determine what bandwidth is needed and how much data is recoverable if a disaster occurs.
Kashya installs an agent on the host systems, but has an appliance that does the actual replication at the primary site. The company claims that this minimizes CPU on the host because most of the work is being done at the appliance level. Kashya also has an appliance at the secondary site.
"In our opinion, their message should be to the companies that have put remote mirroring on the back burner for so long that there is finally an easier and cheaper way to deploy reliable and solid remote mirroring software, versus the old traditional way of doing it on the storage subsystem," Asaro says. "They can enable medium-sized shops to implement a DR where they have none today."
The Right Product at the Right Time
The only trick now is to get noticed.
"I am not sure there is any magic to that other than being there with the right product at the right time," says Mehran Hadipour, Kashya's VP of product marketing.
Hadipour adds that Kashya is working with OEMs, resellers, DR service providers, and consultants to help with market access.
"DR is often looked at as an enterprise-wide problem, and if we assume that everyone has more than one brand of storage, then data replication is not completely captive to top-tier storage vendors," asserts Hadipour. "Customers who put SANs in place hoping to have value other than connectivity would like to have it do more, and fabric-based applications like Kashya is where the SAN promise comes true. Kashya has a number of unique and desirable features that are not available in [existing] storage solutions."
Kashya claims the data replication and backup software segment it is targeting accounts for almost two-thirds of the overall storage software market, citing IDC. IDC predicts that backup and replication software alone will present an almost $5.8 billion opportunity by 2006.
"There is little doubt that cost-effective data protection is the primary challenge facing enterprise IT managers," states Kashya CEO Michael Lewin. "At the same time, they have to deal with the dual constraints of tighter IT budgets and an ever-increasing volume of data at risk that have combined to overwhelm conventional models of data protection. Kashya has adopted a radical new approach to solving the problem that addresses the cost and distance performance degradation issues inherent in existing solutions by creating a more intelligent solution that also lets the user define specific levels of data protection."
Applying Advanced Algorithmic Techniques to Storage and Security Concerns
Kashya was founded by Lewin, VP of Engineering Ziv Kedem, and CTO Yair Heller, who met when they worked in an elite technology unit in the Israeli Defense Forces. Lewin, Kedem, and Heller later earned a reputation within Israel's technology sector as experts in the application of sophisticated algorithmic techniques to complex storage, communication, and security problems. Kashya's management team also includes senior executives recruited from IBM's Storage Group and Zambeel. Battery Ventures and Jerusalum Global Ventures are the company's backers.
Kashya's architecture is based on the intelligent Kashya Data Protection Appliance that connects to the SAN and IP infrastructure, providing bi-directional data replication across any distance for heterogeneous storage and server environments.
According to the company, the KBX4000 appliance is an end-to-end solution for enterprise data replication that leverages advanced algorithms to intelligently manage data and bandwidth, and to deliver guaranteed consistent replicas of business-critical data in the event of any possible failure or disaster, even when that data spans multiple heterogeneous storage and servers.
The appliance also includes a snapshot history that allows rollback to any point in time for quick recovery of critical data. The KBX4000 supports multiple transactional-consistent snapshots at the remote site, allowing reliable recovery in database environments. Frequent, small-aperture snapshots just seconds apart are utilized to minimize the risk of data loss due to data corruption.
Kashya says a key element of the appliance's design is the recognition that all enterprise data is not of equal value and that different data sets require different levels of protection and replication frequency. The KBX4000 offers a full spectrum of replication modes, from synchronous to asynchronous, and from small aperture-snapshot to point-in-time. Users can define different levels of replication based on application requirements and business objectives. Once the parameters are set, the replication process is managed automatically.
Bandwidth reduction technology dramatically reduces WAN costs, particularly over long distances, Kashya says. Data reduction is accomplished through application-aware and storage-aware algorithmic techniques that conserve bandwidth well beyond traditional compression technologies.
Kashya claims it also eliminates the need for expensive edge connect and protocol converter devices and uses existing data centers as secondary sites to reduce the cost of creating a new remote replication site. Storage systems at the primary and secondary sites do not have to be the same, giving the users the option of deploying lower-cost storage or leveraging existing storage.
The KBX4000 will be available in October. The company also provides a global round-the-clock call center and same-day onsite service for hardware.
Back to Enterprise Storage Forum