CloudFoundry: The New Kid On The PaaS Block
CloudFoundry launched this week with a lot of hoohah. It certainly puts it up to Azure and Google App Engine in the PaaS space, but I don’t think Amazon have much to worry about. The ability to build a concrete implementation of a hybrid cloud with the same stack running inside the firewall as outside is the most compelling aspect.
My belief is that the company to crack hybrid clouds will be the big winner in the cloud space and this is the best shot yet (as long as you are using Java, RoR or Node.js).
They didn’t say much about their messaging/queueing offering and I think this is because the integration points between the most popular messaging platforms and CloundFoundry’s supported programming environments is not cleanly defined i.e. can RabbitMQ be used with Java and can can SwiftMQ be used with RoR, more importantly, are they used in this way?
The critical success factor for CloudFoundry will be how easy is it to add new core components. The Open Source approach means that the private cloud guys can hack their own components in, but how easy will it be to get VMWare to allow other components into the mix. Conspicuously absent were, Cassandra, PHP, Python/Django, Oracle and Postgres. Of course anything from Win32 stable was persona non-grata leaving a lot on the field for Microsoft to pick up.
The dirty little secret of course is that CloudFoundry is actually the best invention ever for selling vSphere and vCenter virtualisation to the private cloud gang. Even if, for instance, RedHat were to port CloudFoundry to KVM it would be an implict endorsement of VMWare’s strategy, and RedHat doesn’t like endorsing competitors strategies.
Of course the real vendor in a quandry here is IBM. Late to the private cloud party and desperately rebadging the last decade’s technology as cloud (P690 anyone?). Late to the public cloud party and finding it impossible difficult to sell a retail product (ever watch an elephant try to use chopsticks) . Now blown out of the water by an Open Source offering in a space which IBM would consider its own personal high ground when competing against Azure.
Expect a flurry of acquisitions as the also rans play catch up. Nice dice VMware!
More jobs for those that already have them
Silicon Republic had a great article recently reporting on the explosion of jobs in the tech sector. The IDA chief appears to be postively crowing about his success. Well that’s great, but it won’t move the unemployment needle one basis point.
Why? Because The tech in Ireland sector has virtually 100% employment, take a look at the indeed.ie search for software engineers to see the indigenous requirements.
It takes 4 years to make a new graduate engineer and the current intake of computer scientists and engineers is a fraction of what it was ten years ago. This means we get the same number of technical experts chasing a vastly expanded universe of technical employers.
The last time this happened was in the previous tech boom in 2000. Then, TCD was stamping out over 150 computer scientists a year from various different disciplines (Maths, Computer Science, Engineering). Today I think the total output each year is less than 50. Other colleges has experienced similar downturns.
The net effect of this is a spiralling wage inflation for technical staff, which is good if you are an engineer, but rotten news if you are an employer and an absolute disaster if you are a start-up competing for technical expertise.
Foreign Direct Investment companies (think Google, Oracle, Intel, IBM etc. etc.) who are moving here because it is a “low cost” development centre are in for the same surprise that companies who opened up in India got. Huge velocity of staff between jobs, sign on bonuses and rapidly inflating salaries.
In the noughties we could address shortages by importing expertise from overseas but who wants to move to a country that everyone in the world thinks is an economic basket case?
What to do? First of all understand that bringing in FDI companies that pay practically no tax revenue to the Irish government are of little utility to Ireland inc. from a balance of payments perspective. Second of all they create an unbalanced market as they can continue to afford to pay inflated salaries which squeezes the price up for indigenous companies competing for staff. Those companies are generally in worse shape financially and less able to withstand the stress of these salary increases.
Secondly we need to make more techologists and make them faster. In the 80’s companies like Nixdorf ran conversion courses for the thousands of arts graduates who couldn’t find jobs in our devastated economy and the tech colleges provided a huge array of cross training. Their is a huge opportunity to retrain those that are capable of switching to a technology career (like most career choices its not for everyone).
Pity our training agency FAS is a such a busted flush, we could really do with them right now.
Cloud Vendor Taxonomy
I wrote up a post giving a Cloud Vendor Taxonomy over on the NewBay blog.
Customer Development: From Drills to E-Readers
Customer : I need a drill
Seller: What for?
Customer : To make holes in a wall <<—Pivot: Drills to Holes
Seller: Why do you need holes
Customer: To attach a shelf <<—Pivot: Holes to Storage
Seller: Why do you need a shelf
Customer : To Store Books
Seller: Have you looked at our storage solutions?
Customer: Yes but I like to browse my collections, boxes and cupboards don’t work
Seller: Do you have a lot of books?
Customer: Yes, four or five stand up bookshelves
Seller: What kind of shelf are you attaching?
Customer: Just a single shelf
Seller: How come not another stand up bookshelf?
Customer: I’ve run out of space <<— Pivot: Storage to e-Readers
Seller: Have you looked at our new e-Readers?
Customer: What’s an e-Reader?
Its Movember – Day Zero
My Day Zero Mo.
Microsoft : How many of these mouse are sold globally (at what cost?)
Microsoft's Excessive Packaging for an Entry Level Mouse
I bought this mouse today in town. Got it home and finally managed to break through the packaging to the tiny contents. Not very green.
Cloud Computing Cost Models : Don’t Sweat It
I get asked reasonably often to help companies and individuals come up with a cost model for their cloud computing. People get really exercised about the cost of hundreds of compute nodes and terabytes. I know what these models should look like because I built and insanely complex model for PutPlace in 2006 when we founded the company and decided to deploy it on Amazon. I had the same concerns that most people had, was I building a business that was going to explode in my face because I had made some fundamentally flawed economic assumptions?
Once we launched PutPlace we rapidly discovered a number of interesting facts about our cost structures. The first one was that in a small online business such as PutPlace the compute costs dominate to the point that storage, bandwidth and transaction costs are essentially rounding errors. The second thing we discovered is that attracting enough users to move the needle from a compute perspective is “ahem” challenging for most companies. With consistent upload rates of over 10k files per day (with occasional peaks exceeding 100k files daily) our grid wasn’t even breaking a sweat. We had absolutely no red-line events on compute and of course AWS happily absorbed everything we threw at it without blinking.
Even at the end of 2008 when the service had been up and running for 6 months over 75% of our costs were compute nodes.
So if you want to understand your cost basis a very simple model is to work out the number of compute nodes you want to run, price those nodes in AWS (or Slicehost or Rackspace) and use that as a monthly cost model for your whole environment. Once you have a few months of price data, you can subtract your compute costs to find out your variable costs in storage, transactions and bandwidth (which I’m betting will be marginal). Now you have the marginal costs you can compute your variable cost per active user and now you know what your cost-plus price model basis is for each user.
You should still analyse your bill once a month to prevent surprises (like when we discovered we had 12 months of database snapshots taken at 5 minute intervals that no one was cleaning up) and to understand how the dynamics of your system are changing, but your key focus should be on your overall business model and your customer acquisition strategy.
Think of cloud computing like any other variable cost in your business, when you are small they are marginal (have you every priced up electricity usage in your startup financials?) and if you get big it just becomes a cost of doing business, so don’t sweat it!
(Most of the above applies equally well to building a “scalable” system, NoSQL boosters should take note!)
Going Local vs Going Global when Outsourcing
I did a presentation called “A Cheap Date with Cloud Computing” at the a French Chamber of Commerce event last Thursday .
During the presentation I made an off the cuff comment about staying away from Irish outsourcing companies because they were too expensive. This comment was reported by Chris Horn which elicited some responses online.
Twitter is not a good medium for debate so I thought I would clarify my comments.
- Irish web design/dev shops are more expensive than overseas suppliers and they all seem to get business, so more power to them. This is free market economics at its best.
- Overseas is not just India/China. You can find good, low cost developers much closer to home within two or three timezones of Ireland
- The tech requirements for the audience at The French Chamber of Commerce (primarily SMEs and mostly non tech sector) is limited in scope and doesn’t require high powered (i.e. expensive) design skills or code cutting capability
If you are building genuinely novel software for a new or emerging market using agile techniques then there is some argument for going local because proximity to your development team is a key requirement.
However even in those projects a significant amount of “grunt” work is required (building wordpress themes, connecting payment engines, porting software to different platforms etc.). This work is an ideal candidate for outsourcing and Odesk and Elance do an excellent job of providing a marketplace for lowering the cost of these activities.
As for design work, well I will never ever source startup logos from anywhere other than 99designs.com, end of.
Dublin Startup Weekend @ The NDRC
Oct 1st to 3rd
Don’t forget to book tickets for Dublin Startup Weekend 1st – 3rd Oct @ http://dublin.startupweekend.org/
BOOKING CLOSES 29th SEP
This event is an intense 54-hour event that brings together people with great ideas for web or mobile apps together to develop them from concepts to operational, launch-ready systems over the weekend.
We’re seeking a good mix of developers, designers, bus dev, marketers, and all round innovators, to make this event bigger and better than the last one.
The last event resulted in the incorporation of 3 start-up companies, all still working to develop their products, and everyone who took part (whether they went on to start-up or not) spoke of the value of the weekend.
• To register, visit: http://dublin.startupweekend.org/
• To express your interest, see who else is interested, and receive updates on the event over the coming weeks, visit:http://events.linkedin.com/Startup-Weekend-Dublin/pub/402541
• To read about the event in Silicon Republic, visit: http://www.siliconrepublic.com/start-ups/item/17540
Tickets cost €75 and numbers are limited.
Revahealth.com Changes its Name to Whatclinic.com
I used to slag my buddy Caelen King about the crappy name he chose for his medical tourism search engine RevaHealth.com. Well, he bit the bullet recently and changed the name to the much better whatclinic.com. He also shared some great stats with me,
- Whatclinic.com now has over half a million visitors a month up from 190,000 at the start of the year
- Every day 800 people submit email inquiries to clinics on our site an an additional 900 people call clinics on our site.
- The main treatment categories being looked for are; dental, doctors plastic and beauty
- The average treatment value being looked for €2,000
- We do business in 20 different countries
- UK is now our strongest market and is growing exceptionally well
Whatclinic.com doesn’t just do overseas, it covers Ireland. For instance here are searches for Dublin and Newry.
With Anglo just having dropped another 8.2bn this quarter its great to see a business thriving amidst all the misery and woe.
Here’s a slideshare of Caelen that shows how they go about there business.
