Innovative and enterprise-shifting, these New Data Center wares are tops
1. VMware's Infrastructure 3
Few New Data Center architects need an introduction to VMware server-virtualization software, but some might not be familiar yet with the latest version - VMware Infrastructure 3 (VI3), which offers increased power for virtual machines, upping the symmetric multiprocessor support from two to four processors and increasing the overall memory limit from 4G to 16GB. Additionally, VI3 rounds out storage support with iSCSI and network-attached storage interfaces.
The VirtualCenter management tool also has become significantly better with the new release. It now supports resource-pooling and automated distribution of server and storage resources on the fly. Plus, it now includes a distributed-resource scheduler that load-balances and distributes resources to virtual machines, and VMware HA, a high-availability feature that moves and restarts virtual machines off failed hardware.Longtime user Bruce McMillan, manager of emerging technologies at Solvay Pharmaceuticals in Marietta, Ga., says he especially likes the latter feature. "In the past, if an ESX Server went down, I'd have to go into VirtualCenter and manually start up another one," he says. "With VI3 and HA, it's all automated. That's huge."
2. Tangosol's Coherence
What if you could move and access data as quickly and as easily as you can deploy a virtual machine? That's the idea behind Tangosol's Coherence. The software, which runs on most commodity servers, provides in-memory data management for high-transaction applications, thus eliminating delays that might result from fetching data from a disk or storage area network. Because Coherence runs on multiple servers in a fault-tolerant, clustered fashion, as you add or remove servers, the cluster dynamically reorganizes itself and all the information that it's managing," says Cameron Purdy, Tangosol's CEO. That information is virtualized across those servers, making it available to applications seamlessly as needed.
Wachovia Bank, a financial services firm in Charlotte, N.C., is making good use of Coherence's on-demand capabilities. The bank uses another best product, DataSynapse's FabricServer, to virtualize its Java applications across the enterprise and to broker computing supply and demand, ensuring that each job is performed by the best resource.
"But FabricServer can't leverage the memory of many machines at once, and it doesn't synchronize the data. I need that kind of transactional capability and management, and Coherence solves that problem," says Tony Bishop, director of product management at the bank. With Coherence, performance has improved by as much as 100 times, Bishop says.
3. DataSynapse's FabricServer
DataSynapse's FabricServer is all about matching processing needs to computing resources on the fly. The software provides a control layer that acts as a broker between applications and their computing resources, essentially virtualizing applications so they can be processed by the best available resource. A centralized broker in the software employs user-defined rules to determine that allocation. If an application is deemed time-sensitive and mission-critical, for example, a rule would determine that it be processed only by the fastest, most powerful servers in the data center. Software on each pooled resource communicates with the central server to identify available CPU memory and processing power, says Shayne Higdon, a DataSynapse vice president.
"If I have memory-, CPU- or I/O-intensive stuff, but I know I have a better machine over here that could run it faster than what it's on, I can just allocate it there, since I'm virtualizing the demand and matching it up. FabricServer understands that a service can change its size and shape at different times of the day, week or month, and adjust for that," Wachovia's Bishop says.
"The [Coherence-FabricServer] technology combination has exponential impact in terms of performance and time to market," Bishop says. "Not only do you drive efficiencies, but you can deliver new business capability in a dynamic, real-time manner. It changes the whole paradigm."
4. Newbury Networks' Presence Platform
As much of a blessing as mobility can be, keeping track of personnel and resources and securing the network can be a curse. Newbury tackles the curse head-on with Presence Platform, based on the firm's Location Appliance. The appliance works on any vendor's wireless LAN and tracks any device sporting a WLAN card. It also tracks items more specifically via radio-frequency tags. Newbury provides key applications including location-based content provisioning; or users can use Presence Platform to build custom applications.
"Customers build different rules about what kinds of wireless devices can be active where, and then they can control access to the network based on where users are," says Brian Wangerien, vice president of marketing at Newbury. "So in a hotel, you can provide different access and billing whether you're in a conference room, a lobby or on the concierge floor."
Martin van der Meer is the information adviser at Medisch Centrum Alkmaar (MCA), a hospital in the Netherlands. MCA, which has a Trapeze Networks WLAN, uses the Newbury appliance to blanket its 1,200 users and 200 wireless devices, such as laptops and PDAs. By integrating location information with the Electronic Nurse File patient-tracking system, the hospital makes sure each nurse matches up with the right patient. "With Newbury, the nurses get a short list of patients in our Electronic Nurse File, depending on which room they are in at the time," van der Meer says. A nurse selects a patient and accesses the relevant medical records, rather than having to scan a bar code on a patient wristlet.
5. Incipient's iNSP
Storage virtualization became a lot more interesting last September, when Incipient launched its iNSP software. Running on a storage blade within a Cisco director-level switch, the software offers data migration between heterogeneous and tiered storage environments, all while the arrays are online.
Although storage vendors have been offering similar wares, until now they worked only within their own switches and arrays. By moving the storage intelligence off the array and onto the network in a vendor-agnostic way, and by allowing live data migration, Incipient opens many options for its customers.
Bart McDonough, director of systems architecture at a hedge fund in Stamford, Conn., has been beta-testing iNSP. "Incipient allows us, behind the scenes and in real time, to migrate data over to another array. We can get all the storage off Array A, do maintenance, power it down, do firmware, test it and make sure it's good, all before we start moving the systems back to it. And the production applications aren't affected. It's great," he says.
As much as he likes the high availability, which makes no array a single point of failure, the productivity factor is what sold him, McDonough says. "Bringing down one of our big arrays for maintenance . . . requires a lot of coordination, and a lot of weekend work for my staff. With Incipient, all of this could be done during the workday. That's a huge productivity hit we could avoid."
6. Tideway Systems' Foundation
In large organizations with multiple data centers, inventorying which systems and applications are running where - and how they've changed - can be a monumental task. Tideway Systems' Foundation, a configuration management database-based application discovery-and-mapping tool, provides that information and more on a daily or even hourly basis. The agentless tool finds every application instance - whether known to IT or not - and maps it to any application dependencies on the network, showing users at a glance what's running and what's changed. With that information, users can make proactive, intelligent decisions, knowing in advance the effects of any changes they may make.
Tideway "helps us manage change and improves our ability to spot and analyze problems across the environment," says Stephen Ashton, London CIO at Dresdner Kleinwort, a global investment banking firm. Ashton says he scans his environment weekly but could do it on a daily basis if necessary.
7. Tealeaf Technology's Tealeaf CX
How do you control customers sprinkled all over the Internet and outside your environment, and make sure the performance and service levels they receive meet your expectations? The Linux-based Tealeaf CX network appliance monitors HTTP and Secure-HTTP streams, then "sessionizes" that data, correlating user requests and network responses to build a single-user view of an individual Web session, the company says. IT sets alarms and thresholds, and is notified when users receive site errors. In addition, a "playback" capability lets IT view exactly what a Web customer did and saw throughout a session. This lets Web businesses focus their time on only the biggest, most money-losing errors. Esurance, an automobile insurer, uses Tealeaf in testing new features before they go live, in troubleshooting site problems and in customer service. "Using Tealeaf, we can look up by an e-mail address or different identifiers and pull the session and replay it. It's like a miniature movie. It's invaluable for troubleshooting," says Marj Hutchins, director of Internet operations at the San Francisco firm.
8. Composite Software's Composite Information Server
Whether their legacy databases and systems are ready or not, enterprises are moving to virtualized environments. Composite Information Server (CIS) works to virtualize data queries, letting companies cull data from various repositories without going through lengthy integration or data-warehouse initiatives. A multithreaded Java application, CIS noninvasively accesses data from disparate sources; combines, abstracts and simplifies it; and delivers it via Web services or relational views.
Enterprise Architect Gary Lien uses CIS to provide a single view of members of Life Time Fitness, a health club company in Eden Prairie, Minn. "We can build a single view of data from our cafés, spas and club point-of-sale systems. Without CIS, we would have had to query the data sources and combine the results using a J2EE application, which would have been difficult."
9. Mimosa Systems' NearPoint
Continuous data-protection products, which perform live data backups without affecting performance, and archive the data so it's easy to retrieve, have become a New Data Center staple. NearPoint uses an agentless technology called transaction log-shipping to capture every e-mail, document and file in an Exchange system. Self-service features let users retrieve e-mail without IT intervention. "We looked at several archiving products, and Mimosa was ahead of the game," says Andrew Gahm, network architect for IS at Virtua Health in Marlton, N.J. "Everyone else was using journaling for archiving, which isn't the most effective way to archive your e-mail because it misses so much. Mimosa's log-shipping captures every bit of detail that's in your e-mail, without impacting your Exchange server."
10. A10 Networks' IP-to-ID Service
Just because you know an IP address is generating malicious traffic doesn't mean you can pinpoint the cause of the problem. A10's IP-to-ID Service, working with the company's IDSentrie appliance, eases the identification of network users and devices, and cuts the time it takes to identify and correct problems. IDSentrie drops on a network, establishes links to its directory services, and creates a virtual data store to view those directories. Once a device or user authenticates to a directory, the service takes that directory information and correlates it with the media-access-control address and IP address to develop accurate identity information for the user at that IP address - in real time.
The service is relatively inexpensive, and its simplicity is the real selling factor, says Asad Sarabi, vice president of IT and logistics at Etón in Palo Alto, Calif. "When I see my network is slow or I have a problem, with A10 I can easily, within 10 or 15 minutes max, pinpoint what application, user and device are the cause."
Cummings is a freelance writer in North Andover, Mass. She can be reached at firstname.lastname@example.org.
< Next story: How to avoid bumps on the road to grid computing >