Archive for May, 2007

Dealing with an ever more complicated infrastructure

http://www.sda-asia.com/sda/news/psecom,id,16389,srn,4,nodeid,1,_language,Singapore.html

A recent study done indicates that IT professional worldwide struggle with the increasing complexity of data center management while at the same time facing sever budget and personnel constraints. And to ease management challenges, they are relying on several different technologies and initiatives including storage capacity management, virtualization and Information Technology Infrastructure Library (ITIL) frameworks.

The survey — The State of the Data Center study– is a two-part research series highlighting top trends and challenges for data center managers. Conducted in April 2007, the first part of this study included more than 500 respondents from enterprise-class companies in the North American, European and Asia Pacific markets.

Conducted by security company Symantec Corp the responses of the study indicates that the growing volume of servers, applications and operating systems in today’s data centers is contributing substantially to data center complexity.

Check out this article, its talking about a data center study by Symantec which talks about the growth in server volumes, applications and operating systems, how the IT infrastructure is getting more complex, even though I might virtualize the operating system, I still need to administer and support it, how we deal with this will be the issue. The problem being that for each operating system or suite of operating systems I might have one management tool, that there still isn’t one platform through which I can easily manage the different virtualization platforms, Microsoft, ESX, Virtual Iron and Xen Source, these things will come, in the meantime, speaking with people to see what they’ve done seems the way forward, an interesting article, check it out.

HP and infiniband for C-Class blades

http://home.businesswire.com/portal/site/google/index.jsp?ndmViewId=news_view&newsId=20070524005295&newsLang=en

ALO ALTO, Calif.–(BUSINESS WIRE)–HP (NYSE:HPQ) today announced two agreements with Cisco that target the fast-growing market of high-performance computing with HP hardware and Cisco’s InfiniBand solutions.

First, HP will offer Cisco double data rate (DDR) InfiniBand-based Server Fabric Switches (SFS) as part of HP’s Unified Cluster Portfolio. Additionally, HP BladeSystem c-Class servers are now supported with Cisco’s standards-based InfiniBand host driver software.

The combined offerings accelerate business growth for high-performance computing (HPC) environments by providing high bandwidth, low latency, fabric stability and scalability.

HPC configurations are used for projects requiring millions of calculations running simultaneously, such as financial data analysis, data mining, market data analysis, computational chemistry, database tiers, and computational fluid dynamics. Cluster computing is one of the drivers of the HPC market because of its flexibility and price-performance advantages.

HP is to offer Infiniband functionality for its C-Class blades, very cool, this will bring new functionality to the platform as well as new opportunities for their use either in a grid or high performance computing solution. Check it out.

How many data centers should I have?

http://www.techworld.com/opsys/features/index.cfm?featureid=3394&pagtype=samechan

Despite the consolidation trend, some companies are continuing to branch out with their data centres for reasons including redundancy of the centralised data centre for backup, a more comprehensive disaster recovery plan and better uptime at regional offices.

Indeed, some companies actually prefer this approach, and experts say it makes sense for organisations such as those with branch employees who can act as IT technicians or firms with locations dispersed on different continents.

Deciding how and where to deploy your data centers is an important strategic decision from a cost and a business strategy process viewpoint, having everything  in one location might expose you to unnecessary risk, at the same time, having many different data centers for each location or business line might introduce unnecessary cost, you need to balance the costs/risks as with anything, the interesting thing is how virtualization will change this - will it allow my IT to follow the sun with the business? To have my Singapore users use my London servers when the electricity is cheaper and I have people on site?

ClearCube gives users control over virtualized desktops

http://www.eweek.com/article2/0,1895,2136258,00.asp

When ClearCube launched the updated 5.5 version of its Sentral management software April 9, the Austin, Texas, company announced that it was supporting VMware’s ESX Server 3 along with VMware Server virtualization software.

Now, starting May 24, ClearCube is looking to offer a blade-based computing system that will support VMware’s VDI (Virtual Desktop Infrastructure)—software that allows users to deploy virtualization technology across an enterprise’s entire fleet of PCs—which will allow users to support between 12 and 15 virtual machines on each physical blade.

ClearCube have been working on blade pcs for years, and they continue to innovate, the article discusses how using the ClearCube software and VMWare, the user can not create virtualized desktops - how cool is this? The ability to give the end user control (with the right controls in place) could bring real benefits in deployment, empowering the IT to meet the business need - will check this out, sounds very cool.

IBM makes the data center a more efficient proposition

http://hardware.silicon.com/servers/0,39024843,39167294,00.htm

IBM has unveiled a data centre which it claims uses up to 15 per cent less power than a traditional data centre.

The modular data centre can be designed in eight to 12 weeks with 210 blade servers sitting in each rack.

John Moseley, IBM’s UK site and facilities business manager, said the steel infrastructure of the data centre comes in “flat pack boxes” and can be put up anywhere.

Check out this article about IBM’s example of its setup to help reduce data center power by up to 15%, anything any of the vendors can do to assist with energy utilization in the data center has to be a good thing, do check it out, its an interesting read and has some pictures - very cool.

Data center requirements an ongoing challenge

http://www.echannelline.com/usa/story.cfm?item=22041

A recent survey conducted by Aperture Research Institute of 100 enterprise data managers representing over 600 data centers, indicates that more than 43 per cent of respondents reported that 90 per cent or more of their data centers were in use.

Check out this post which is talking about the demand for data center space, it covers some important topics, and its always good to read what others are thinking about and it covers some interesting topics, very cool.

Patching Apple OS X again

http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9021259&intsrc=hm_list

May 24, 2007 (Computerworld) — Apple Inc. today unveiled the year’s fifth major security update for Mac OS X to patch 17 vulnerabilities, but fewer than one-third of them could lead to hackers injecting their own code into a compromised system.

It appears it’s time to patch my Mac again, have done my duty, clicked the update button and rebooted it, in the meantime Computerworld have done a document about it explaining all the patches etc.

How many hotspots does your data center have?

http://www.computerworld.com/action/article.do?command=viewArticleBasic&taxonomyId=154&articleId=293918&intsrc=hm_topic

May 28, 2007 (Computerworld) — When Roger Hardy, IT director for the city of Jeffersonville, Ind., gets an alert from an automated monitoring system that his data center air conditioning is failing, he has 20 minutes to fix the system before the computer room’s temperature reaches what he describes as its “death point.”

If the thermometer inside the data center hits 91 degrees, Jeffersonville’s IT equipment is cooked — literally. That’s what happened this month when the city lost $20,000 worth of equipment after its strained air conditioning system shut down during a spell of warm weather.

Check out this article it’s talking about how cooling capacity can be delivered adequately but you still need to examine the air flow, to try and minimize the number of ‘hotspots’ you have, to aid airflow in cooling the data center as much as you can. With blade servers this is particularly important, ensuring that the air flow is consistent that the air gets sucked through the blade not allowing it to settle around the cabinet creating a ‘hot spot’, cooking the blades, do check you’d be surprised how much difference an extra grated floor tile can make.

Memory will drive the market forward?

http://www.echannelline.com/usa/story.cfm?item=22044

Server vendors will continue to be winners but the microprocessor vendors will see their sales decline with greater server consolidation occurring among end users. That is the conclusion of the latest server report from IDC, said Matt Eastwood, program vice president of enterprise platforms at IDC.

Even with server sales down to a 4.9 per cent worldwide growth in the first quarter of 2007 and previous 10 to 15 per cent growth patterns for these machines a thing of the past, vendors like Hewlett-Packard, Sun, IBM and Dell are still better equipped to find new revenue sources with virtual machines, courtesy of the customer demand for increased memory, I/O sub systems and storage, as well as software and services, he explained.

The article raises some interesting topics about virtualization and memory being a key opportunity in the market. It’s right in some respects ensuring that you have adequate memory for the virtual machines is important. Check out the article, it covers some interesting information about the virtualization market, you never know when they might be needed for your next presentation.

Chargeback remains a topic of interest virtually speaking

http://weblog.infoworld.com/virtualization/archives/2007/05/capacity_and_ch.html

Before virtualization came around, IT departments typically had a process in place to account for department chargebacks when it came time for billing departments for computer resource consumption. Now, many of those systems just don’t make a whole lot of sense.

Check out this article, it’s talking about how virtualization has been affecting chargeback in the virtual world - it remains a hot topic, how to do it strategically, should you do it? What products to use one of the pay for use ones or an open source one, should it be linked into the other way you cost your support etc, by business line, do check it out, it covers some relevant points.