Devoted to hardware, software, IT issues, and information technology management

I understand that this is probably old news, but I just finished reading the press release from Royal Dutch Shell plc (NYSE: RDS-B) regarding their efforts to outsource a significant part of its IT infrastructure & telecommunications services to AT&T, T-Systems and EDS.  I find that I need to vent a little bit as this will impact approximately 3,000 IT staff worldwide:

"The move is part of the company's initiative to achieve top quartile performance in its businesses and functions in support of its "More Upstream, Profitable Downstream" strategy.

As well as significant improvements in efficiency and productivity, the initiative will deliver important financial benefits for Shell over the first five years.  Over the same period, the total value of the Agreements to the three suppliers will be in excess of $4.0 billion."

While they expect to transfer the majority of their IT staff to the service providers, there will assuredly be a reduction in headcount...there always is.

Sorry for the month-long hiatus, but I'm back at it again.  I ran across an interesting post on The Register that described a software solution from ScaleMP that will allow a customer to link two dual-socket x86 servers via InfiniBand and have them behave as if they were a single four-socket server.

Why not just buy a single four-socket server you ask?  Well, according to ScaleMP there are two compelling reasons:

• Two dual-socket servers are less expensive "to purchase" than a four-socket server (TCO aside...).
• Hardware vendors have a tendency to introduce more advanced and faster processors in their dual-socket servers first.

As usual, The Register is a regular fount of knowledge when it comes to all things IT.  I was reading an article and learned that the SANS (SysAdmin, Audit, Network, Security) Institute, reportedly one of the largest sources for information security training and certification in the world, also maintains a site called the Internet Storm Center (ISC):

"The ISC was created in 2001 following the successful detection, analysis, and widespread warning of the Li0n worm.  Today, the ISC provides a free analysis and warning service to thousands of Internet users and organizations, and is actively working with Internet Service Providers to fight back against the most malicious attackers."

Well, I thought we had addressed all of our issues related to the initial change of Daylight Savings Time (DST) back in 2007.  Server and Client operating systems were patched; so were messaging environments and anything supporting Network Time Protocol (NTP) was looked reviewed.  It was even necessary to make some changes to systems that synchronize e-mail to mobile handsets such as RIM's BlackBerry Enterprise Server or Motorola's Good Mobile Messaging and patch the handsets themselves.

Preparing for DST three weeks early and one week later seemed almost reminiscent of the buildup to Y2K.  We didn't experience too many glitches other than known problems with Outlook calendars and I thought we would be set after making it through 2007.  Low and behold, DST part deux has come along and flummoxed quite a few folks.

When it comes to hardware failures in PCs, servers or storage arrays, I've always operated with the knowledge that any component with moving parts is the component most likely to fail and the faster the component moves, the lesser the mean time between failure and more likely that it will fail.  That said, it's no revelation that hard drives, spinning at speeds of 5400, 7200 or 10000 RPM, are going to fail; which is why RAID and global hot spare drives should always be used to automatically respond to a failure and prevent the loss of data.

Increasing the availability of a storage array usually includes providing other redundant components such as controllers, fans, physical data paths, power supplies, etc., but again the disks were always assumed to be the most prone to failure.  Well, I just finished reading an article in The Register titled "You don't know disk about storage failures" that discussed a report compiled by the University of Illinois Department of Computer Science and Network Appliance illustrating that while disk failure contributes to the majority of storage subsystem failures, other hardware/software should not be ignored when purchasing a storage array.

A few years ago I read an article in Popular Science that discussed the creation of a team within the military devoted to guarding against and responding to cyber attacks.  At the time, we were just starting to see intrusions and data breaches so sophisticated that there was a lot of talk about state-sponsored hacking similar to state-sponsored terrorism.  What really caught my attention in the article was that protection and detection weren't the only focus; countermeasures and response tactics were being hinted at.

Well, according to a recent article on The Register, the U.S. Air Force has a new Cyber Command which is scheduled to be operational in October and they've just released their strategic vision for the new unit:

"Controlling cyberspace is the prerequisite to effective operations across all strategic and operational domains--securing freedom from attack and freedom to attack.  We will develop and implement plans for maturing and expanding cyberspace operations as an Air Force core competency.  We will provide decision makers with flexible options to deter, deny, disrupt, deceive, dissuade, and defeat adversaries through a variety of destructive and non-destructive, and lethal and non-lethal means.  Finally, we will do this in friendly cooperation with our professional partners and teammates in other major commands, Services, combatant commands, and U.S. government agencies."

This is old news from the end of February, but I've only just gotten around to dropping a post on it.  Novell Inc. (NASDAQ: NOVL) has acquired PlateSpin Ltd., a leader in server virtualization and data center management tools.  Novell, best known for the NetWare operating system and top-of-line directory service has lost a lot of traction in the data center to the arguably inferior Windows operating system, as well as UNIX & LINUX and they've been flying under the radar (at least my radar) for quite a while now.

I think this is a great move by Novell to make a splash in the virtualization arena because PlateSpin has a few great products out there:

• Forge is an appliance solution (think hotsite in a box) that provides complete protection for up to 25 server workloads straight out of the box.  In the event of production server downtime or a disaster, protected workloads can be rapidly recovered and continue to run on the PlateSpin Forge appliance until they can be restored to the production environment.
• PowerConvert provides the flexibility to move and rebalance workloads in any direction between physical and virtual hosts.
• PowerRecon is a feature-rich analysis tool that can identify targets for consolidation or virtualization and you can convert IT resource usage data into billable values for chargeback purposes.

About a week ago I posted Google Storing Medical Records and discussed this emerging service.  At the time I was concerned that the medical information Google is storing would not be governed by the same HIPAA regulations that ensure the privacy of this data.  I was also concerned that Google could sell this data to biotechnology or pharmaceutical companies who would then use it for direct marketing purposes.

One key source of revenue that I completely forgot is ad revenue and now I've just read a post on the Bits blog in the technology section of the New York Times that reminded me of this potential gold mine.  Google's chief executive, Eric Schmidt, stated that "at first there would be no ads on the service", but a Google spokesman also stated that "the company may well put ads on future versions of the service".

If you've read my biography and some of my posts, you'll know that I'm an almost rabid proponent of server virtualization and since I cut my teeth on VMware ESX Server, I'm a huge fan of their products.  I like to know that I'm getting the most out of my expensive, high-performance hardware and virtualization allows me to do just that.

Another important aspect to virtualization is the level of isolation and security provided to the virtual machine (VM) and the server or desktop hosting those VMs.  This helps to ensure that a misbehaving application won't adversely affect the function of the VMs operating alongside it or the underlying host.

In recent years this feature has provide protection for end-users by allowing them to use VMware Player or Workstation to create a VM on their desktop that is solely dedicated for use when accessing the Internet.  Should the VM ever become subject to a virus, spyware or other malware; rather than trying to extricate the malware (sometimes almost impossible), the user could dispose of the infected VM and deploy a fresh one.

Last December I posted Where In The World Is AMD's Barcelona? to highlight the supply problems AMD was experiencing in delivering their quad-core processor to the marketplace.  One of the main reasons for the shortage is that most of the available processors were being shipped to directly to Texas to be installed in a high-performance computing (HPC) system called Ranger, which was expected to become the fastest supercomputer in the world with a peak performance of 504 teraflops when it entered service on February 4th.

There were several delays, not all attributed to AMD, and after IBM recently updated their HPC system (Blue Gene) to support a peak performance of 596 teraflops; the Texas Advanced Computer Center (TACC) can only claim Ranger as the fastest "open system" HPC in the world.  Ranger is the largest computing system in the world for open science research.