One of my first postings was concerning virtual server technology and its industry leader, VMware. I am an avid user and supporter of VMware’s product and use it extensively in my employer’s network. Like many veterans in the field, I feel that virtualization is where the future is headed in IT.
Since that posting, VMware has been the victim of a roller coaster ride. The stock price last year shot up from its IPO in the upper 20’s to 125 a share only months later. Industry journals and magazines were continuing a steady stream of positive articles on the new virtual giant.
Since then, things haven’t been as rosy for the ten year old firm. While nearly all stocks have fallen, VMware’s stock price plummeted to $38. Its parent company, EMC, fired its CEO who led the company since its inception. These symptoms stem from the fact that a giant killer has its sights set on them, Microsoft. The software giant released its own hypervisor (the industry term for a virtual server manager) in July. Like the Internet Browser in the 80’s, Microsoft realizes that it nearly missed the boat on the virtualization market. Like most its first generation products, its hypervisor is not near the product that VMware has.
VMware’s Vmotion provides a degree of high availability for its virtual server that can’t be matched by any of its competitors. VMware also runs more efficiently and has a lower I/O rate than Microsoft. Can you sense a “But” coming. Here is the “But”. Microsoft has their hypervisor priced at only $28 when appended to the purchase of Windows Server 2008. VMware’s basic ESX hypervisor is priced at $495 per dual-core processor. $2,995 for its full package that includes Vmotion.
Though their product is superior, it is obvious that the company will not be able to sustain its pricing model in a growing competitive market. History also shows that Microsoft usually unveils a fully competitive product by its third generation so VMware is most likely on the clock. Microsoft has killed some mighty competitors in the past such as Netscape and Novel. The question that virtual advocates are asking is what will be VMware’s next move that will allow them to escape the same fate.
Thanks to Orin Optiglot for permission to use this Photo.
I was reading CSO magazine the other day (Corporate Security Officer) and came across a fascinating new security product for executives who transport highly confidential company documents. A pharmaceutical issued a high security briefcase for its research executives. The briefcase has two security modes. The “Loss Proof” function alerts the executive with an alarm signal when he is more than five meters away from the case. This is to not only discourage someone from lifting the case, but prevents the owner from leaving the case incidentally unattended.
The “Robbery Proof” mode is designed so that if the owner is accosted and forced to give up the case, the case will wait until the briefcase is more than 100 meters away and then send a 30,000 volt shock throughout the case as well as sound an ear shattering alarm. Click here to check out a similar case.
With Apple continuing to dominate the headlines and industry buzz with one acclaimed release after another, it’s easy to forget that the so-called ”Evil Empire” has been rather quietly refining their Surface computing technology. We’re talking about more than multi-touch features for your mobile phone; Surface is perhaps most intriguing for its’ “surface” adaptability and potential virtual applications. Check out this video demonstration of Microsoft’s Sphere prototype. I’m an Apple guy myself, but I must admit that Sphere could be interesting…
Ryan Rode
Interactive Services Manager
Ashworth University
Thanks to commorancy for permission to use this Photo.
In an earlier post, I mentioned the use of SAN storage (storage area network). A SAN is basically a bank of very fast hard drives. Servers can then access the SAN for file storage through a separate dedicated network, allowing data to flow between the SAN and the servers swiftly and securely. SANs allow for centralized control over an organization’s data. Traditionally, network administrators relied on DAT, direct attached storage, in which the file system for a server resided on the server’s local hard drive system, usually a RAID 5. This highly complicated the back up process since the data was spread out amongst different hardware. It also offered a poor utilization of hard drive space as some servers were always short on free space and some servers had lots of space that was never used. A SAN allows an organization to maximize its hard drive utilization and makes backing up the file system a snap. Two of the biggest players in the SAN market are EMC and Netapp.
SANs are very expensive however. A single 300 Gig drive currently costs somewhere around $2500, not to mention the rest of the SAN hardware. A new company called RevStor has an interesting SAN alternative. It is marketing a product call SANware Mesh, an application which installs an agent on every computer in an organization, looking for available hard drive space. This includes both PCs and servers. The application then distributes stores the data for the organization across the network on the available hard drive space via the agents, which divide, encrypt and distribute the data. No more than 33% of any file is stored on a single computer. Every file is also copied onto separate locations in order to enforce redundancy. SANware Mesh therefore maximizes the DAT drives throughout a network. The application also allows the administrator to centralize the backups.
The primary downside of this storage model is that unlike a SAN, the data is transported across the primary network, rather than a separate private network as a SAN does. This diminishes the performance speed of the network as network traffic must compete with data traffic. SANware offers considerable savings over a SAN and thus is a viable alternative for small organizations. For more information on this interesting product, go to www.revstor.com
Thanks to tifotter for permission to use this Photo.
YouTube enthusiasts were denied access to their daily dose of online videos earlier this year when the site went down for two hours back in February. The culprit was not due to any type of hardware failure, but due to the direct efforts of the Pakistani government. Pakistan, like a number of countries, such as China, Thailand and Turkey to name a few—regularly attempt to monitor internet traffic by their citizens and block designated sites, especially YouTube.
Pakistan Telecom, and the country’s main ISP PCCW, blocked access by hijacking YouTube web address. Anyone based in the country who then attempted to visit the video sharing site was redirected to an unknown alternative site. Unfortunately, the hijack didn’t stop there, and was leaked to the wider Internet.
This meant that ISPs around the world started blocking access too, which resulted in the site being unobtainable by users in Germany, China, USA, Russia, the UK, and Australia. The problem lasted for approximately two hours, before YouTube engineers issued a statement concerning the problem, and PCCW stopped the blockage.
The event has proven to be more than just an embarrassing mishap for the Pakistani government, but has brought up concerns as to the vulnerability of the Internet. Is it now possible for a country to bring down designated website traffic world wide?
The management of updates has become an essential element in overseeing a computer network today. Take the network that I manage for the school district I work for. Being a Windows based network, our 90 servers and 5500 workstations rely on Microsoft update which are made available from Microsoft every second Tuesday of the month for the most part. Critical updates are made available on an as needed basis such as a hot fix or the plugging of a security hole that has been exposed.
When I first entered the networking field, these updates had to be managed on an individual computer basis much like your home computer. This of course was very labor intensive. It also ate up a lot of your Internet bandwidth as each and every machine had to communicate directly to the Microsoft website. Today, most windows based networks of any size utilize a Windows WSUS server, (Windows Software Update Services) which download s all available updates and then dispenses them throughout the network during off peak hours. See the link for more information: http://technet.microsoft.com/en-us/wsus/bb466193.aspx
The most common types of updates are for antivirus and anti spam applications. Ensuring that these types of updates are downloaded and installed throughout the network is of the utmost importance. Viruses can spread across the world in a matter of hours and can bring down critical servers or even an entire network unless AV applications are current.
Like most schools and many other types of organizations we have a web filter. Nearly every night, the filter downloads the latest site category updates. To explain, a filter company constantly discovers new sites across the Internet for categorization and will return to previously visited sites to discern if the site needs to be recategorized.
We have dozens of applications such as our backup software that regularly download updates as well. In addition, all of our servers are from Dell, which periodically (about once a quarter) sends out a CD with hardware and BIOS updates for its machines.
Thanks to David Cardoso for permission to use this Photo.
Ten years ago the big upgrade in network speed was from 10 MB to 100 MB. Organizations quickly began upgrading their LAN infrastructures to take advantage of this increase in performance. This meant of course that CAT 3 cable had to be replaced with CAT 5 cable and all the switches and routers that couldn’t handle 100 MB had to be replaced. Most organizations had to upgrade the NICs in all of their computers as well. Five years ago, organizations started upgrading their datacenter structure to 1 Gig for all of their servers in order to enlarge the pipeline of data moving into and out of the datacenter from users.
Although some organizations have provided 1 Gig throughput for their entire network, meaning that once again most switches, routers and NICs have to be replaced, many have not felt that the return on investment (ROI) is worth it. Despite this, vendors are now touting the new 10 Gig Ethernet standard. It will be interesting to see what type of demand for a speed that will most definitely require an organization to upgrade all involved switches and routers and NICS. The cabling for 10 Gig is different as well.
There are not many people who can seriously put into question the projected dominance of Google search technology in the future evolution of the Internet. Through his brilliantly engineered software architecture and conceptual maps of its' potential applications, supergenius Tim Berners-Lee, the acknowledged inventor of the World Wide Web, has for years been carrying his message of a "semantic web" that will make the current incarnation of the web seem like your Dad's Rolodex by comparison.
In Berners-Lee vision, the next generation of web technology will be semantic, i.e. all data will be interconnected and capable of communicating with other "information" through a common language so speak. The following article discusses the development of the semantic web alongside the related, but not parallel development of Google’s search, mapping, and tracking applications. Berners-Lee is not anti-Google, but he is passionately against the centralization of web data and any form of proprietary control over Internet content.
I also encourage you to check out this very insightful video presentation in which Berners-Lee takes you on a virtual tour of the Web from its’ beginnings into tomorrow. This is fascinating stuff. Please share your thoughts with the Ashworth Blog community after brainstorming on your own.
Thanks to Dana Lee for permission to use this Photo.
Have you seen the recent commercials from Verizon Wireless touting their wireless service, contrasting their service with the seemingly restrictive confinement of WIFI. This is part of a battle for wireless dominance. WIFI is the current wireless standard leader. If you have a laptop, chances are you have some sort of WIFI router device in your home to provide WIFI access for it and other devices in your house such as the newest IPOD release. You have probably utilized the WIFI offerings of a local coffee shop or hotel (I couldn’t imagine staying in a hotel today that didn’t offer WIFI or at least Ethernet network access). If you download a free program called NetStumbler and install it on a computer that has a wireless card, you can discover wireless access points wherever you are. The program makes a sound every time a new WIFI point is discovered. Driving around with your laptop running NetStumbler can make your computer sound like a video game as it will sound off repeatedly like a machine gun as it find the plethora of WIFE points in the area.
The problem with WIFI is that it is spotty and not predictable. WIFI has a limited range and many WIFI hotspots are now guarded with security such as WPA, disallowing public access. This has created a demand for a more streamlined wireless standard that offers blanket coverage.
One option is the cell phone companies such as Verizon. These services can offer as much as much as 3.5 Gig of network throughput, putting WIFI to shame. WIMAX is another option in which metropolitan areas can offer blanket coverage throughout the area with speeds up to 4 Gig. Both of these offerings work on some sort of subscription plan which can cost as much as $60 a month. Unlike WIFI, each and every device must have a separate subscription, making these alternatives very expensive. Over a dozen metropolitan WIMAX projects have been dismantled over the past year due to higher than predicted costs and a shortage of willing subscribers. WIFI has also fired back with the new 802.11n WIFI standard this year which offers greater throughput and coverage than its WIFI predecessors. The battle has just begun for wireless dominance…
If you are purchasing a PC today over $700, chances are the specs include a dual core processor. Multi-core processors have been thetrend for several years. A dual core processor is a CPU with twoseparate cores on the same die, each with its own cache. It’s the equivalent of getting two microprocessors in one. A typical server today has at least two dual core processors. Quad cores are also very popular too. The organization that I work for typically uses dual quad core processors for any of the servers that will be carrying a processing load such as a database server, email server or a virtual host.
But as you can surmise, it’s not stopping at quad core. Intel will be releasing a six core processor later this year. AMD will be releasing theirs early next year. These will be strictly targeted at the server market since desktop software can barely fully utilize dual core processing. And if 6 isn’t enough, both companies plan on releasing a 12 core processor sometime in 2010.
Subscribe to Feed
