Bladewatch.com

http://www.bloomenergy.com/products/what-is-an-energy-server/

Built with our patented solid oxide fuel cell technology, Bloom’s Energy Server™ is a new class of distributed power generator, producing clean, reliable, affordable electricity at the customer site.

Fuel cells are devices that convert fuel into electricity through a clean electro-chemical process rather than dirty combustion. They are like batteries except that they always run. Our particular type of fuel cell technology is different than legacy “hydrogen” fuel cells in four main ways:

I was doing some research about power in the data center and came across a few articles talking about the Bloom Energy Server, it’s a power generation technology based on fuel cells, further research noted it being deployed by several organizations for mixed uses. It will be interesting to read more about the solution and examples of it being deployed, I wonder if it might be a vehicle to power data centers, and also a way to make your energy costs more fixed or scalable, you buy just what you need and grow your fuel cell installation in line with your anticipated energy requirements. Even if it was not something that you wanted to commit to powering your data center directly, could it be a UPS alternative, or something that you deploy along side your grid or hpc solutions rather than keep adding data center power capacity and load? We’ll have to see, I’ll need to read up more about it.

So in reference to this post, we’re looking at four major products as part of the bladewatch research activities:

Veeam – backup solution for VMware

Dell PowerEdge C6100 server for cloud builders and high density environments.

VKernel vOPS Reporting and chargeback – something that I know has been a topic of interest for colleagues and our blog

HP Proliant DL980 G7 – the 8 socket server powered by AMD processors – what does this platform have to offer for today’s requirements and is there demand for the 8 socket server in the x86/x64 space?

I got asked by Chris how this would work. Well basically, we go off to the product pages, read all about them, phone a few friends and ask their views, then put together some words and analysis.

Want to feature in our weekly four products/services of the week?

Email us: martin237@gmail.com, but please put in the subject “Four products/services of the week”, Martin gets about 200 emails a day in press releases, announcements and questions.

There is no ulterior motive, it’s simply a way of keeping up to date, we don’t ask that you be a multi national but we do ask that you take a look at the blog and decide if your product or service would be relevant to our readers.

We’re off to look at Veeam, it’s something that has been getting rather a following in the virtual space in the backup arena and as a production support guy, who knows the importance about backups, it’s always interesting to see what alternatives there are to the traditional backup solutions that are in the marketplace. I’ve been hearing good things about Veeam from several colleagues, in the meantime, to see what I’m talking about, here’s a bit from the Veeam.com web site:

http://www.veeam.com/vmware-esx-backup.html

Veeam Backup & Replication is built specifically for VMware vSphere to provide fast backup and recovery for ESX(i) virtual machines. Only Veeam Backup & Replication lets you recover an entire VM or an individual object from any application or file system—all from the same image-level backup. There are no agents to manage, no need to babysit backup jobs, and for the first time you can verify the recoverability of every backup.

It’s no wonder Veeam Backup & Replication v5 was chosen as “Best of Show” at VMworld 2010 and why more than 20,000 customers use Veeam to protect their mission-critical VMs.

I was interviewing an IT Manager for an article as we sometimes do over at Bladewatch and I was asking their view on their virtualization project. It’s interesting to see what others experiences are and to read olden days posts where we have been talking about virtualization of servers to say a four socket server like a HP Proliant DL585, DL580 or the IBM/Dell and other equivalents. As time moved on we started talking about how some people were using blades and NetApp storage for their virtualization projects, the blades due to volume, the NetApp thanks to its excellent storage management and features like roll back, thin provisioning etc.

It was during this interview that it reminded me about a conversation I was having with a colleague, is it the bad boy large 4 or 8 socket servers that you want for virtualization these days or the mid range dual socket machines which have enough network ports and PCI type slots for storage or extra network connectivity?  Interestingly my friend was going on that you want Dell R910s, or HP DL580s or either boxes AMD equivalent, why? The volume of virtual machines he could host on one of those more powerful machines.  I wondered and was asking but wouldn’t you rather have certainly for the majority of the estate, more smaller ESX or ESXi servers as they were typically cheaper to buy and therefore easier to sign off, and when a machine had to be taken down for maintenance it was less impacting?  I got told that I was some what missing the capacity point, I suppose it’s one of those customer and site specific debates, but I still feel either a blade or a midrange 2u server like a Dell R710, the IBM x3850 would be ideal, twenty or thirty of those in bulk zoned into the storage, with the networks presented gives you more flexibility and would typically give you enough time to spread around load whilst you get the next set of infrastructure signed off.

For the second time, I was talking to Chris, who continues to try and invite me over for lunch in Sunny Canary Wharf. He was having some what of a quandary. A user had phoned up and started making a series of requests before he put in his P2V dvd and pressed go, to transform the application server in this case, an IBM x3650 to a virtual machine, a virtual instance. The application manager sent him a variant of the email below, we’ve removed anything that could identify Chris, but the context, the essence remains the same.

Danny,

We’ve got the network connections ready for the ESX host, and can therefore commence virtualization of server18763 when you are ready, we’re looking to do this online due to the nature of the services and applications involved, however we can do this offline if you prefer.

Can you please confirm the date and time so that I can raise the change.

Thanks

Chris Smith

Windows Analyst

Global Technology

The Application Manager (or Danny’s response)

Chris,

That’s fine but can you please confirm the following:

When you virtualize the machine, what kind of RAID do we get, at the moment we have RAID 1+0 for the OS, and RAID 5 for the data.

What adjustments do you need to make to the backups and will this change the nature and window allocated to this task

How will virtualization of the machine affect our disaster recovery procedures and will be mirroring the instance so that if live goes down, service continues?

Cheers

Danny

Application Manager

Global Services

So the debate Chris and I were having over the phone was one of context and process. Chris was asking if high availability was part of the virtualization project, his task to plug in P2V dvd, press virtualize seemed to differ from re-architecting the global services application. Also he was not aware of the disk set up as it was on SAN or NAS storage and that was handled by the storage team, in terms of backups the backups ran as they always did, in terms of window and performance he did not anticipate change in schedules but then that was managed by the backup and storage teams.

My reply to Chris was to respond something like this – it was off the top of my head, so would end up being rewritten several times before it was sent out:

1. The nature of the machine is changed, the server is virtualized into a container which is hosted on a high availability cluster of ESX servers, with the virtual disks residing on high available disks with replication where possible and where the application or infrastructure mandated that it was required.  Further discussion regarding disk configurations would need to be discussed in partnership with the storage teams.

2. The nature of the backups remains the same, however as part of the change the team responsible for the backups will check the configuration and ensure that it conforms to our operational standards and any deviation escalated to the relevant business sponsors

3. The virtualization project aims to transform the physical server to a virtual instance which can be hosted on our new virtualization cluster with in built resilience and scalability, however in order to facilitate additional or enhanced functionality in terms of disaster recovery we need to understand the application and infrastructure requirements as well as their constraints in order to provide the right platform going forward.

If I’ve missed anything, or you have any thoughts, send me an email or post a comment.

http://www.finextra.com/News/Fullstory.aspx?newsitemid=22580

The poll of more than 100 CIOs and other IT executives shows 76% plan to increase technology spending over the next 12 months, 17% are cutting investment and seven per cent keeping levels steady.

• Understand what we have
• Stabilize the core = make what we have (server/network/storage/application/monitoring) work better – optimize configuration, stabilize configurations and reduce outages – fix disk/memory issues, rebuild where necessary, Keep The Show On The Road
• Consolidation of the data center
• Server consolidation and virtualization including (or) hardware refresh
• Converged infrastructure
• Drive towards energy efficiency
• Reduce complexities and improve resilience and a move towards that zen goal of dynamic infrastructure

Irrespective of cloud, of service level agreements and contracts, a simple question one which might seem strange.  Your service provider says it has full fault tolerance, disaster recovery, a series of business continuity and reliability processes to ensure that it can meet its contractual obligations and maintain service for your business and your customers.

My questions?

What do you mean? A stand alone pc which you can re-ip?

Business continuity or disaster recovery you have said that’s included but what is the actual downtime involved and what is the anticipated timeline to recovery.

It’s great to see cloud services, outsourced solutions creating opportunity for all from a customer, supplier and service provider standpoint, but I wonder if as the costs come down we don’t detract functionality, or create a gap between what users perceive to be in place so that in event of a systems failure, service continues.  To the small service provider I met many months ago who told me, “oh our DR is that we switch the production server off, re-ip the DR box, and restore data from tape”. In many respects that is providing the ability to continue service, but at what operational cost both financially and in terms of service, sorry your site is out of date but we’ll copy the pictures over soon….

http://www.finextra.com/News/Fullstory.aspx?newsitemid=22579

Mizuho’s ATM network and online services were knocked out in March, leaving millions of customers unable to access their money in the wake of the earthquake and subsequent tsunami.

The problems were caused by a surge in donations made through Mizuho’s ATMs following the disaster and the bank’s systems then struggled over the next few days to cope with the backlog of unprocessed transactions.

Check out this article talking about organizational change in this bank in an effort to improve the reliability and scalability of it’s IT systems.

http://www.oracle.com/us/corporate/press/401816

Insurers struggle with business agility and see their legacy IT systems as a hurdle to optimizing customer service and faster time to market for new products and channels, according to a study released today by Oracle Insurance and Insurance Networking News.

The study of 421 insurance industry professionals – including C-level and senior executives, IT professionals and line of business managers – looked at organizational priorities for insurers, their effectiveness in priority areas, and the role that IT plays in helping and hindering their initiatives.

An article from Oracle talking about their research into what insurance executives want from their IT and how IT can be perceived as being a hindrance to achieving these goals. It is and it isn’t. Firstly IT as an individual and a department needs to understand the rules of engagement and discover how it can add value, in some cases that may be doing the support, the architecture or the vendor/service provider negotiation and management, once it understands it’s roll only then can it act as an empowerment tool, “you’ve asked for this, have you considered…”

Change can be uncomfortable but it’s part of life, and it’s easy to sit on the sidelines and pass judgment or make comments, IT and service delivery really is a two way street. The business need to manage and explain their expectations whilst understanding that there is a cost of doing business, can we run your department on a desktop with a an inkjet printer, yes, will it have the same resilience and performance as an enterprise solution no, so on that basis where is the middle ground, where can I identify savings and offset the right costs to deliver what you want on time, on budget and quite possibly with the flexibility in the solution to deliver on demand, real or perceived.

http://www.bladewatch.com/images/virtualization.pdf

I got asked by a colleague earlier on in the week if I had ever completed the book that I had started about virtualization, I’d made the mistake of mentioning it to them asking for their opinion as to what it should include.

Anyway, I did write one and it remains on my trusted rather olden days original Mac Pro. It is not particularly long, but felt that the pace of change currently going through the industry combined with understanding what to cover from a technical standpoint could make it a complicated challenge, it would it be one of those things that I wrote and then looking back as I do at my posts think “awwwh look what I was working on, thinking and worrying about”.

With that, I have left online for all to see and freely available the virtualization slide show that I had put together many years ago with a viewpoint of covering virtualization from an end user perspective, I know a few recruitment guys and some users had seemed quite happy with it.  The document remains unlocked, it remains unbranded so you can carry it around, re-use it or show it to people without massive logos, just mention bladewatch if you get a chance.

The document covers business concepts and illustrates how you should maybe go about adopting it as a concept or platform in your business though I recognize everyone is different, there is no fancy technical speak, no technical server specifications all business like, all understandable.