HP's ProLiant Gen8 servers have a self-diagnostic system that collects 1,600 different data points on a server's operation. HP thinks it would be better to use that information in preventing the server from going down, instead of fixing it afterward.
On Tuesday HP unveiled Always On Support Services, its redesigned technical support for systems living in a more virtualized world. Virtualized pools of data center resources call for a different form of support than the sort that's sufficed in the past, according to HP.
The embedded diagnostic system can feed its data into a piece of HP monitoring software designed to scrutinize it around the clock, watching for anomalies that mean trouble. If that data emerges, the Insight Remote Support software can analyze the issue and notify support personnel in an HP call center, such as its Atlanta Mission Critical Support Center. Technical staff members take over from there, notifying the server owner of an impending problem and advising them on what to do about it.
The support service also works with older generations of hardware from HP and other vendors, but they produce less system data to work with.
HP's Flynn Maloy, director of support marketing, says this type of tech support is actually cheaper to provide than the kind that fixes things after they're broken. He said HP will make skilled specialists available at any time of day to deal with technical issues, and it's HP's goal to have the new style of support supplied by someone who already knows your company.
[ Want to learn more about HP's converged infrastructure? See HP Upgrades ProLiant Servers For Converged Data Centers. ]
The new program in some ways parallels the Oracle Premier support that it introduced with its Exalogic and Exadata appliances. The appliances feed data back to Oracle diagnostic systems, with technical support specialists calling customers before the customers know they've got a problem.
The two programs illustrate the increasing importance of technical support as an intrinsic part of operations. "We don't believe virtualization has reached its full maturity," said HP's Maloy. With the increasing complexity of converged infrastructure, where the server may be closely linked to network switches, some problems have the potential to cascade into other systems if allowed to develop too far.
When they do, "there's no smoking gun telling you what's wrong. You don't get a red light at the top of the server that went down," he said. What's needed is consistent monitoring and correlation of anomalies with a large database of knowledge about similar running systems. Such systems can connect the dots when the temperature rises, the response time slows, and I/O ports back up.
Too often, technical support consists of "ninjas who parachute in to get it up and running again" after a server has failed, Maloy said. HP will extend the data collecting capabilities found on its latest ProLiant servers into other parts of its converged infrastructure, including storage and network components.
HP's new support comes in three classes:
--Foundation Care, where a customer gets a contact at HP who will leverage HP's relationship with software vendors to identify and fix problems.
--Proactive Care, aimed at minimizing downtime by correcting problems before failures occur, with direct access to HP experts in the problem area.
--Datacenter Care, customized support for a customer's multi-vendor environment, with use of predictive analytics.
Maloy wasn't able to supply an example of how the preventive care would prove less costly to the customer than reactive care; pricing will be on a case-by-case basis, he said. The Always On Support will be a featured topic at HP's Discover customer meeting, scheduled for June 4-7 in Las Vegas.
IT's spending as much as ever on disaster recovery, despite advances in virtualization and cloud techniques. It's time to break free. Download our Disaster Recovery Disaster supplement now. (Free registration required.)