A Storm of Servers: How the Leap Second Led Facebook to DCIM

Last July 1, that scenario became real as the “Leap Second” bug caused many Linux servers to get stuck in a loop, endlessly checking the date and time. At the Internet’s busiest data centers, power usage almost instantly spiked by megawatts, stress-testing the facility’s power load and the user’s capacity planning.

via A Storm of Servers: How the Leap Second Led Facebook to DCIM.

What was happening? The additional second caused particular problems for Linux systems that use the Network Time Protocol (NTP) to synchronize their systems with atomic clocks. The leap second caused these systems to believe that time had “expired,” triggering a loop condition in which the system endlessly sought to check the date, spiking CPU usage and power draw.

How Facebook threatens HP, Cisco, and more with its “vanity free” servers

Facebook, Amazon, and Google are all very picky about their server hardware, and these tech giants mostly build it themselves from commodity components. Frank Frankovsky, VP of hardware design and supply chain operations at Facebook, was instrumental in launching the Open Compute Project because he saw the waste in big cloud players reinventing things they could share. Frankovsky felt that bringing the open-source approach Facebook has followed for software to the hardware side could save the company and others millions—both in direct hardware costs and in maintenance and power costs.

via How Facebook threatens HP, Cisco, and more with its “vanity free” servers | Ars Technica.

The OCP hardware designs are “open” at a higher level. This way anyone can use standards-based components to create the motherboards, the chassis, the rack-mountings, the racks, and the other components that make up servers.

Facebook Saves Datacenter Costs with Frigid Arctic Wind

The constant, biting wind may have stunted the growth of Lulea’s tourism industry, but it has proven a big factor in luring big IT facilities into the area. Datacenters in Lulea are just as difficult to power and cool as any other concentrated mass of IT equipment, but their owners can slash the cost of cooling all those servers and storage units simply by opening a window: the temperature in Lulea hasn’t stayed at or above 86 degrees Fahrenheit for 24 hours since 1961 (PDF), and the average temperature is a bracing 29.6 Fahrenheit.

via Facebook Saves Datacenter Costs with Frigid Arctic Wind.

Get An Inside Look At The Rackspace ‘Castle’ Via Google Street View

Recently, we had a photographer come out to create an inside Google Street View of our HQ so our customers, friends and family – anyone, really – can get a virtual feel for what it is like to work at Rackspace. There’s a ton to see.

via Get An Inside Look At The Rackspace ‘Castle’ Via Google Street View – The Official Rackspace Blog.

HP Calls Out Cisco With Data-Center Switches

The star of HP’s show, or at least the product with the biggest number, is the FlexFabric 12900 core switch, which can fit 768 10Gbit/s ports or 256 40Gbit/s ports. Cisco’s 18-slot Nexus 7018 claims to have the same 10Gbit/s density but only has cards to support 96 40Gbit/s ports.

via Light Reading – HP Calls Out Cisco With Data-Center Switches.

Calxeda’s ARM server tested

At first sight, the relatively low performance per core of ARM CPUs seems like a bad match for servers. The dominant CPU in the server market is without doubt Intel’s Xeon. The success of the Xeon family is largely rooted in its excellent single-threaded (or per core) performance at moderate power levels (70-95W). Combine this exceptional single-threaded performance with a decent core count and you get good performance in almost any kind of application. Economies of scale and the resulting price levels are also very important, but the server market has been more than willing to pay a little extra if the response times are lower and the energy bills moderate.

via AnandTech | Calxeda’s ARM server tested.

As usual another thorough review from Anandtech.  Below is another interesting architectural tidbit.

CalxedaSoc_575px

Let’s start with a familiar block on the SoC (black): the external I/O controller. The chip has a SATA 2.0 controller capable of 3Gb/s, a General Purpose Media Controller (GPMC) providing SD and eMMC access, a PCIe controller, and an Ethernet controller providing up to 10Gbit speeds. PCIe connectivity cannot be used in this system, but Calxeda can make custom designs of the “motherboard” to let customers attach PCIe cards if requested.

VMware targets rival “bookseller” Amazon with its own public cloud

The public cloud confirmed today by VMware marks the first time the company will become an IaaS provider itself, analysts said. Unlike platform-as-a-service, which puts the focus on providing simplified tools to application developers, IaaS clouds let customers (or require them to) manage the underlying infrastructure such as the operating system and virtualization tools.

via VMware targets rival “bookseller” Amazon with its own public cloud | Ars Technica.

The Leading Open Source Data Center Automation and Configuration Management Tool

Puppet Labs’ IT automation software enables system administrators to deliver the operational agility and efficiency of cloud computing at enterprise-class service levels, scaling from handfuls of nodes on-premise to tens of thousands in the cloud. Puppet powers thousands of companies, including Twitter, Yelp, eBay, Zynga, JP Morgan Chase, Bank of America, Google, Disney, Citrix, Oracle, and Viacom.

via Puppet Labs: The Leading Open Source Data Center Automation and Configuration Management Tool | Puppet Labs.