System Administration

HP is pushing intelligence and sensors end-to-end in their newest generation of Proliant servers.  The intelligence gained from their Sea of Sensors is critical to proactively predict and reactively respond to hardware conditions and causes of these conditions.  By logging and analyzing the data from these sensors, HP is hoping to correlate events or conditions in one area that may cause a fault or failure in another.  And by aggregating this data all up into their Insight Online cloud offering, they hope to aid customers by using the data gained from other customers.

I spoke with John Gromala, director of product marketing for industry standard servers at HP, and Keith McCauliffe, chief technology officer for industry standard servers at HP, about the innovations and features that were built into the Proliant Gen8 in the bloggers lounge.

By building all the hardware sensors and notifications directly into iLO, the system is online and monitoring the environment and hardware before the POST process completes on the main system.  iLO also has its own logging built-in and the optional Insight Online will allow the information to centrally logged for analysis by the customer and HP, enabling proactive alerts and proactive service calls to prevent failures.

“Not only are we monitoring the information, we’re logging the information persistently, said McCauliffe.  “In the event of a server fault, what were the pre-existing conditions before that happened?”

“Sometimes you don’t even know what changed,” Gromala added.

That sort of information is invaluable to getting the system back online as quickly as possible.  As with backup, its not fast backup, but fast recovery that the user really cares about.  Identifying what failed or had a fault at the hardware level may be critical to resolving the issue quickly.

“As system designers, our engineers are trying to look for cause and effect, proper behavior or misbehavior,” said McCauliffe.  “We’re going to put a lot of sensors in because we want the telemetry of what’s going on, but we want useful tools to abstract that.”

And those tools are what make the information useful to end-users.  The tools built into the iLO will offer context to the data the sensors are providing.  And since iLO is really a separate computer on its own, even in the main server will not boot and you cannot collect diagnostic data, the data in iLO would be available for analysis.

Out of all the conversation, I was struck by how responsive many of the innovations in Gen8 are to address real-life problems of the IT administrator.  There are many more innovations in the Proliant Gen8 line that are direct responses to customer requests – such as the enhanced drive carrier with indicators to show when not to remove the drive, a new Service Pack for Proliant (SPP) firmware and driver distribution.

Particularly with the SPP, HP directly addresses complaints of many server administrators that firmware and driver updates are very disruptive and hard to deploy in their production environments.  The SPP offers a way to manage and push updates, many of which no longer need reboots, and stage them in a way that they are activated by a scheduled reboot at a later time.  The center of this innovation is a new version 5 of the Smart Update Manager.  This tool allows an administrator to download a repository of firmware and drivers, which were developed and tested together (another change), and then scan and patch multiple servers from a single console.

The message that this sends to customers is pretty clear, though.  HP is listening to customers and their issues and trying to address them in the best ways possible.  That a good thing to hear as an HP customer, like myself.

Disclosure: HP sponsored my trip to HP Discover, covering all travel expenses, however HP does not control the content of the posts that I write based on my attendance to the show.  The thoughts, impressions and information contained in this post are my own.