On Being Wrong

 Cloud  Comments Off
Oct 182010

I like conversations with people who challenge my views on things. If I’m right then I should be able to marshal my arguments well enough, and if I’m not then clearly I need to learn and change my ideas.

So this evening I had a long chat with Barb Goldworm and she made me think long and hard about a couple of things.

First, InfiniBand. I kind of wrote this off in the same bucket as FCoE because in my view storage is moving towards TCP/IP, and the problems with IP connected storage are probably easier to solve than the problems of switching to yet another new technology. But is that right? Taken to its logical conclusion, every drive would have an embedded NIC and would connect to a backbone with a storage processor directing IO to the disks in RAID and other survivability configurations, or potentially servers directing IO direct to individual disks if advanced features are not required. Logically, this would probably work better with InfiniBand than with IP and quite likely with better performance, scalability and reliability. I’m thinking here about the day when comms between the storage processor (already often an Intel-based server appliance at least in midrange) and the storage itself evolves away from fibre. Fibre has a lot going for it but it is expensive and arcane; “everyone understands IP” in the way that “everyone understands Windows” (including the downsides such as naive and half-assed designs). Could InfiniBand really be the Next Big Thing? Or is it, as I had previously supposed given its decade and more of gestation, a niche technology outpaced by the more agile and more cheaply scalable IP? I will have to look harder at this.

The Microsoft analogy brings me to the second thing. I said already that VMware might become Novell 2.0; I am more convinced of this every time I hear that Paul Maritz is opposed to embracing other hypervisors and doubly so every time I have the conversation with anyone outside VMware, since everyone I’ve spoken to outside VMware basically agrees with me on this. What’s the implication,though, for my favourite hypervisor? Logically if we can choose to use KVM, Xen, ESXi or HyperV, won’t the Microsoft shops all vote for HyperV, or at least do as I would like to do with my infrastructure and split the load etween hypervisors? In my case I want RHEV / KVM on some hosts for operational (read: licensing) benefits, but given the pervasive nature of Windows and the fact that I already by DataCenter Edition licenses for my ESX hosts to cover the guests, surely if vCenter and its attendant (and usually very good) ecosystem of plugins and extensions also covered HyperV, why would I use ESXi instead? I think ESXi is the best hypervisor today but I have not benchmarked 4.1 against R2 SP1 – I am told that the improved memory management in SP1 now places it on an equal footing. Maybe it does.And everybody understands Windows, right? (No, very wrong, actually, ask any Windows techie to install R2 Server Core and manage it, see if they really understand Windows, but you get the idea).

So, some food for thought there and time to shift my opinion on some things, perhaps. Just when I thought I understood it, too. Oh well.


 Cloud  Comments Off
Sep 252010

A shout-out to DreamHost. And to @JoeBaguley of Quest, who has caused me to think harder about the environmental impact of daily life.

Recently while I was away at VMworld we had a circuit breaker trip and for one reason or another the power to Marvin, my server, did not get restored. The impact was immediately visible: a reduction by 0.5kW in power draw. That’s a lot of power for a private website and mail server, so I cast around for a cheap and easy-to-use hosting deal. I settled on DreamHost because the price was right (under $10 per month), I saw some decent reviews, and it has an awesome one-click install process for the software I already used (MediaWiki and WordPress) and integrates with GMail for mail.

I needed some support during setup, the quality and speed of this was awesome. Humbling, actually, I am nowhere near that good with my internal customers. Sure, not quite up to VMware standards but I pay for platinum support with VMware and I can upgrade to a an almost equal level with DH for another ten bucks a month. Those folks are smart and friendly. I can’t fault the value of that offer.

Importing my Wiki to the test site took very little time and setting up the managed DNS for transferring the live site took hardly any longer. I can save the hosting cost in month 1 because I can move to a cheaper ISP which does not provide a static IP, the power saving is free collateral benefit.

Why did I not eat my own dogfood and go for cloud provision, I hear you ask? Easy: I need a completely predictable bill, just as my business units do. So managed virtual hosting works for me as it works for my business and funnily enough for my business’ customers as well. Cloud is about where you stop caring, the monthly bill is something I care about.Blimey, the phrase eating your own dogfood was popularised in Microsoft by Paul Maritz, now CEO of VMware. Small world.

Oh, do feel free to point out where I am wrong, as I am sure I am – no doubt there are cloud providers who are as easy to use, as cheap and as competent as DreamHost. Right now I am in no hurry to change, I have been surprised and delighted at every stage, and the fact that you are seeing this shows that it Just Works. I am all for point solutions that Just Work.


 Cloud, Vendors  Comments Off
Sep 252010

Exalogic: 2010 going on 1985.

It seems I am not alone in failing to see the purported benefit of Oracle‘s megalithic Exalogic platform. Leaving aside1 for a moment the months of FUD Oracle spouted about cloud while they were not ready and frantically playing catch-up, the whole idea of “cloud in a box” seems to me to be missing something pretty fundamental about the idea of cloud, in as much as it has any real meaning. Especially when the box is that big and that expensive.

Larry Ellison said, at the launch event, that Exalogic provides “elasticity”. How? How can a megalithic architecture possibly provide elasticity? The product may well be incredibly powerful and a really great fit for some Oracle-backed, Java-fronted transactional systems, but surely what they have actually done is simply to reinvent the mainframe? A big, expensive, proprietary, professionally supported platform for virtual machines. That sounds uncannily like the IBM 3090 I used as a student in 1985!

I’m equally unconvinced by the Vblock (sorry Steve), and for similar though not identical reasons. At least the VCE partnership don’t claim to be something they are not, but the idea of premium priced preconfigured virtual infrastructures seems unlikely to me to get much traction in the market. We already know how to build that. It’s only the more conservative who will find a compelling case for what amounts to not much more than a validated and supported configuration of standard components that we can buy off the shelf – especially when you consider that not all of these components are necessarily best of breed, and that will become increasingly true over the lifecycle of the product.

For me, these things are un-cloud, maybe even anti-cloud. Cloud is about moving away from vendor lock-in, big iron, the whole business of having one foot nailed to the floor.

I see Exalogic as reinventing something we already knew how to build a quarter of a century ago and decided to do something else instead. I could be wrong about that, but I don’t think so, not this time.

1 Well, no, not really otherwise I’d not mention it, but you know what I mean.

Feb 172010

How many internal clouds have a proper DR mechanism, a recovery environment, regular testing, a mechanism for retiring and reinstating inactive VMs and so on? And at what point in the transition from toe-in-the-water desktop virtualisation to betting the farm on cloud do we start to consider these things? We are mature users of virtualisation and I am still struggling to get some people to grasp the fact that tens of terabytes of data and hundreds of VMs is something for which you might want a continuity plan of some sort. Luckily these conversations are getting much less common as the profile of internal BCP rises.

The last mile

 Cloud  Comments Off
Feb 092010

A great point from yesterday’s #PoweredByCloud : what about the last mile?  I’ve already spent a long time talking to storage companies about the difference between the US and rest-of-world when it comes to dark fibre for data replication, this also feeds into the whole cloud concept. What happens if your customers are on dodgy broadband links, or in Bratislava or Jo’burg or somewhere in China where bandwidth is sucky and unreliable?  For some of us there is an option to fall back to 3G but as @ianhf points out that is not going to work when the outage is caused by a civil contingency, which will also affect the on-ramps for 3G. @JoeBaguley made a good case for “design for failure”, this is definitely part of it.