Servers aren’t designed to be exciting. They’re made to chug along quietly, doing their jobs with no fuss. But when they do go down, Murphy’s Law dictates that it’s at the worst possible moment, and often without warning.
It wasn’t always that way. More years ago than I want to remember, we got a call from our minicomputer’s service tech, saying he was on his way with a new disc drive. We were puzzled; we hadn’t asked for service, and nothing appeared to be misbehaving.
“Oh,” he said. “The computer called and said a drive was failing.” And because the computer monitored itself, we were spared unscheduled downtime.
Fast-forward a few years. Intel servers are now the norm, and I don’t remember one of them ever calling anyone to say it wasn’t feeling well.
But that’s all about to change with the launch of HP’s Generation 8 (Gen8) ProLiant servers.
With this move, HP is addressing the challenges of companies with limited or no IT staff. And, according to Tarun Bhasin, research manager at IDC Canada, it is not the only company doing so.
“All the major server vendors are working toward solving the same customer problems – be it improving energy efficiency, reducing complexity or improving manageability,” he notes. “But the difference is in how they approach these problems. And HP is using these design innovations in an effort to differentiate its x86 server offering.”
The result of $300-million and two years’ worth of research and development, HP says the Gen8 are the world’s most self-sufficient servers. The company claims onboard automation triples administrator productivity in two ways: by eliminating routine tasks, such as checking logs, and by speeding server deployments with embedded tools.
The keys to all this are what HP calls its 3D Sea of Sensors and its Active Health technology. Basically, it has placed components in its systems that monitor more than 1,600 parameters. Therefore, like my minicomputer, the server can let administrators know something is failing before it breaks. It can even notify HP support, or a partner of choice that maintains the machine, letting them know both what is wrong and what parts are required.
Customers and partners can track server health via the HP Insight Online secure portal. If they subscribe to the forthcoming HP Insight Remote Support 7.0 system, they’ll find a picture of their servers’ status, and any pending alerts and service calls. The system even auto-generates support tickets when something needs attention, just as our old minicomputer did.
The portal also provides warranty and service contract information, as well as comprehensive reporting.
HP Authorized Channel Partners who support multiple customers are given that all-important “single pane of glass,” allowing them to keep an eye on all of their support customers’ environments at once.
When the servers are housed in the newly announced intelligent racks, containing circuitry that lets them sense what servers are mounted in them, administrators can easily monitor and control power consumption in their datacenters from a central portal.
“HP has hit the nail on the head with its new features,” says Mr. Bhasin. “Our research indicates that a majority of IT staff time is actually spent on deployment, monitoring and maintenance. And one key aspect of the new Gen8 servers is automation, which helps reduce deployment time and improve energy efficiency while decreasing both planned and unplanned downtime.”
Hardware health is only part of the equation. The controlling software (“firmware”) embedded in that hardware also needs attention, yet is often forgotten. To help keep those components up to date, HP also announced Smart Update, which manages updates to firmware, drivers and tools. Customers can opt to create their own repositories of these items (useful, for example, if certain software or peripheral hardware is not compatible with the newest version), or use HP’s cloud-based updates.
A third cause of failures is – let’s face it – user error. For example, it’s all too easy for an administrator trying to replace a failed drive in an array to pull out the wrong disc. While (so far, at least) the system doesn’t reach out and slap that errant hand away, Gen8 servers have a “do not remove” indicator on active drives.