Bladewatch.com

IT Jungle have a post going over the server market announcement last week, it was an interesting read, the server market is changing in terms of volume and requirements. At the same time I had a few thoughts in the server space:

• Not everyone is at the same point in their IT infrastructure
  • We need to continue onboarding new users, new markets, establishing their unique requirements and what technologies can be levelled to deliver their business benefit
• Virtualization means fewer better specified servers
  • Virtualization can mean more infrastructure spend – once we see our customer specific benefits we might achieve further investment
• We need to take the technology out of IT and into the business world
  • Hardware monitoring is generic IT tool, unless you provide the business specific viewpoint – which means little business relevance and buy in.

Share and Enjoy

http://www.bladenetwork.net/?pageid=1236

PARIS, France, November 25, 2009 – Emulex Corporation (NYSE:ELX) and BLADE Network Technologies Inc. (BLADE), the trusted leader in data center networking, announced that the two companies are delivering the essential networking components for IBM BladeCenter Virtual Fabric. BLADE’s BNT Virtual Fabric 10G Switch Module combined with Emulex Virtual Fabric Adapter CFFh for IBM BladeCenter provide an end-to-end converged networking solution for IBM BladeCenter H and HS22 blade servers.

Virtual Fabric for IBM BladeCenter provides flexible Ethernet connectivity, virtual port switching and support for future converged networking with iSCSI or FCoE. Unlike the physical port switching employed by conventional switch fabrics, the Emulex-BLADE solution provides a virtual fabric for IBM BladeCenter supporting 1Gb/s or 10Gb/s that can be configured into up to eight virtual ports with bandwidth allocation in 100Mb/s increments. 

Furthering the possibilities of converged network connectivity, combining storage and network down the one pipe has to be a good thing, and this platform can bring real benefits in reducing the complexity and the delivery times in getting things done. I’ll need to read up more, well done to Emulex and BLADE Network technologies.

Share and Enjoy

I have written an essay talking about thinking about how we teach IT going forward, what we need to be including how we need to evolve the way we teach like we do in the way we support and provision our infrastructure.

I will post it nearer the end of the week and make it freely downloadable, its a few ideas and resulted from a conversation I was having with dad actually, you see I went to college/university, came out with my degree looking for work and wish that I had been supplied with a bit more ammunition:

• How to behave – suit/no suit – how to respond to those emails, what building a server meant, did I need a screw driver, what a bank build was, those kind of things
• What was best practice – what the vendor said and what we actually did, why we were sometimes a few generations behind, how I could make a difference
• Some technical and non technical experience – to be not armed that sounds emotional, but had I been on call, I might have understood what it meant, what the process might be

I wonder if we could not transform the perception of IT through education, and reach out to the next generation, get them enthusiastic about the platforms, understand their new ideas and as a person who is now 29 – avoid getting stuck in the:

• we’ll see
•  in the we don’t do it that way
• we can’t change things

To get back to how it used to be :

• We can do things better,
• Why are we doing it like that?
• We can make a difference

Something which every IT person has to believe as part of good delivery, development and self belief/satisfaction.

This essay is short, it is not directly helpful for those trying to configure an ILO or deploy a virtualization solution, but indulge me, see what you think and hopefully refresh your ideas, your concepts and continue the IT journey through innovative, excitement and deliverying empowerment to the end users.

Share and Enjoy

http://www.finextra.com/news/fullstory.aspx?newsitemid=20873

Industry association TM Forum says the ECBC has been established to help “understand the needs of the largest global cloud buyers and ensure any impediments to the uptake of cloud technology are removed”.

Other members include Alcatel-Lucent, Amdocs, AT&T, BT, CA, Cisco, EMC, HP, IBM, Microsoft, Nokia Siemens Networks, Telecom Italia and Telstra.

The council has identified several areas they will work on, including common cloud services product definitions, security issues, interoperability, data portability, APIs, service provider benchmarking and network performance and latency issues.

This is great news, reaching out to understand individual end user groups can be a great way of onboarding new users, and in innovating the offerings not only in terms of technology and process, but in billing and integration from an end user standpoint?

Share and Enjoy

We are seeing more vendors come on stream offering cloud based services to assist end users achieve their business goals, Fujitsu is one of these, do check out this url talking about it on eweek.

The more we expand the possibilities and the offerings, the more we can find the right range of cloud services for our businesses, excellent news, off to check out more.

Share and Enjoy

I asked one of the IT Managers to give me some quick tips on how to reduce the carbon footprint of your data center, his replies are below: (Thanks Danny)

• Establish the infrastructure requirements – does this data center, and the underlying applications require N-1 resilience? What tier is the data center and therefore what level of backup and resilience do we need, are we over provisioining?
• Look at the data center temperature in terms of raising it to reduce the cooling costs yet not risk a specific risk to production or a resultant increase in support costs?
• Examine the air flow, the power distribution and the energy efficiency in the data center space
• Use the technologies that are available, lights out, KVM, reduce all the anciliary components that we do not need, whilst managing the on site need at 3am when an engineer needs console access.
• Right size infrastructure, can we ensure that the systems we deploy are not over specified for the actual role, that DL380 necessary for that file server, or could we use a 360? A virtual machine? Could we use SAN storage or NAS rather than an array shelf of physical external disks?
• Consolidate applications, roles and infrastructure – do we need everything that we have, this includes reducing the complexity of the feeds and services – what duplicate roles and feeds do we have in this space – operationally and technically – could we have one service providing the feeds rather than eight different ones owned by several business lines

Share and Enjoy

1. What technologies are you going to be using to deploy the operating system and are those technologies included with the blade servers themselves
2. How do you plan to use these blades with your existing environment, integrated or non integrated switches, 10GB Ethernet or normal Ethernet to work with your existing infrastructure
3. Think about the power load and cooling required in terms of density, how this affects your data center air flow and power
4. Managing servers in volume, managing 4 rack servers is one thing, managing 48 blades is another, in terms of asset management, rebuilds and support – what procedures need put in place, enclosure management etc. As we move to dynamic provisioning, where I can rebuild a server in 30 minutes, how do we manage the needs of the end user to have control with our own internal billing and systems management requirements?

Share and Enjoy

Hi Martin,

Can you summarize a few points about creating an energy efficient blade solution, what would you look at?

Regards

Naresh

Hi Naresh,

Great to hear from you, off the top of my head, if we wanted maximum efficiency, I would be thinking about:

• Right size the infrastructure for the application – this can be the processor, the network, storage and memory, or something as simple as choosing the mid range processor rather than the super fast one to gain the maximum benefit in performance per watt.
• Look at san boot or solid state disks, so we reduce the power/heat generated from the individual blades, also could we have a single image which blades boot from and a temporary write disk, rather than 600 copies of Windows, application and configuration files?
• Integrated network switches to reduce cabling and power consumption – what can we do on the network side to improve not only the efficiency of the network, but the performance of it, can we optimize the code, optimize the infrastructure configuration to achieve the most, using the right firmware, the right switches and vendor recommended configurations.
• Look at the power technologies, can the blades be configured to power down when not in use, can we have energy efficient power supplies, and the own vendor specific tools to improve efficiency out of the box
• Look at data center design, can we use fresh air cooling combined with more effecient air flow and distribution to ensure that we are not wasting effort on in effective heat spots.
  Look at the data center operations in terms of reducing the number of kvms installed, working more in lights out, can we be looking at DC power? Will that help with our energy costs? What differences can we make to the operating temperature, would 25 degrees Celcius not give us reduced cooling costs and have a marginal cost on my hardware support costs – even more so if it is a grid environment, if a grid engine were to fail, do we care?

Share and Enjoy

http://www.cnplus.co.uk/news/business/new-free-air-cooling-technology-used-at-30m-teesside-centre/5211362.article

IT services supplier EDS is using Flakt Woods technology at its new £30 million data centre in Teesside to help the company to leverage “free-air” cooling.

The technology will help EDS save thousands of pounds a year by chilling the air within its data centre in Wynyard near Billlingham on the Teesside business park.

The new centre will be the first of its kind in the country to use this new state-of-the-art cooling method.

It is the size of four football pitches, and will house sensitive data – on 7,000 computers.

Check out this post illustrating how EDS have used free air cooling in this data center to help reduce the temperature of the data center, whilst at the same time saving thousands of pounds in energy savings, its technologies and best practice like this that can transform the possibilities and the cost of the data center. We hear all the time from clients who just need to extend the life of their data center that little bit more for that project, whilst so and so is decommissioned etc, the more we can do this, the more we can be responsive to business need, managing expectations and the cost.

Share and Enjoy

http://www.itjungle.com/tfh/tfh120709-story08.html

The idea of putting servers, storage, and networking gear into metal shipping containers and linking them together into a data center cluster is not a new idea–Sun Microsystems was the first to propose the idea back in October 2006–but it is catching on enough that IBM is endorsing the concept and shipping a product.

Big Blue was showing off its riff on the containerized data center, which is called the Portable Modular Data Center, at the 28th annual Gartner Data Center Conference, in Las Vegas, last week. And as you can see, it is a shipping container with a paint job–well, at least on the outside.

Its great to see IBM offering the data center in a container, that the choice and innovation in this sector continues has to be a good thing, it will be interesting to see the specifications and more information, this blog had some great detail, do check it out.

Share and Enjoy