Microsoft, Oracle finally get a room: Now they're ready to take on Amazon cloud The rise of Amazon Web Services's hulking infrastructure cloud has turned former rivals Microsoft and Oracle into chums. Redmond announced on Thursday that Oracle Database, Oracle WebLogic Server, and Java were all now available on the Windows Azure cloud, served up in license-included virtual machine images, just as El Reg …
Amazon has a very real image problem they need to get sorted. This 'Mackrel' (it's all I could get for a name, sorry) doesn't really have to be a technological equivalent to Amazon's offerings. All it has to be is not Amazon.
There's nothing 'wrong' with Amazon's cloud that isn't a function of it being a cloud. It just doesn't appeal to a lot of technology oriented businesses. You can tell someone that Netflix uses the Amazon cloud and they'll tell you how they can see why that works, Amazon is a bookstore and Netflix is a video rental shop, maybe brick and mortar bookstores should have tried that before they all died.
But Oracle and Microsoft, that's how businesses operate. You don't rent your computing capacity from a bookstore. Don't think for a second big, successful companies are above that sort of thinking. They most certainly are not. Billions of (currency) are spent every year simply because 'Buying (x) could undermine staff confidence, so let's go with (y)'. That's common, that's legal and 'activist investors' won't touch that with a 5,000m pole. It's too much of an emotionally loaded thing and it looks like those investors are attacking general staff, and that never works.
It's all stupid bullshit, but nobody equates (or should at any rate) making lots of money with intligence. Intelligence is not required to successfully run a company, emotion always factors in and if they tell you it doesn't they're lying. The Mackrel partnership will create a large number of excited customers, just because of who they are.
Why would any Oracle products technologists who are not dopey use Microsoft Azure Cloud when Oracle supports and is a vendor of OpenStack Cloud Computing Services like "most every other technology company" in existence except Microsoft? Plus, Oracle products and services have proven, and been widely advertised as working more efficiently with better performance on Oracle's own Linux and Solaris OS and Virtualization/ OpenStack Cloud Computing Services.
The related question is, why would any entity that is a strong Microsoft Cloud and Virtualization proponent and user - meaning the full Microsoft stack - be using an extensive array of Oracle development technologies in the first place?
This strange relation between Microsoft and Oracle offers a convoluted solution to an almost non-existing problem.
"Most companies don't write all their own software so even if they like the MS stack they may buy a product which requires Oracle. Happens all the time."
This. The company I work for has just bought a £30 million a year turnover company that uses Oracle on Windows as its main software offering - the customers who use it by and large have no real idea of whats 'under the hood' as it just works for them.
Now we have to get it into the cloud. Fun and games to ensue
MS do support Linux in Azure, what makes you think that they haven't just put in an appropriately tuned Oracle/Linux stack?
Actually, I've got two hypervisors in my home lab, one running VMware one Hyper-v (with open stack to follow) and the Linux support for Hyper-v is really pretty good, to the extent that you can back up a Linux VM effectively agentlessly in MS' DPM just using the Linux integration components.
Because OpenStack is a horrifically complex and horrible to use pile of crap in comparison at the moment.
And as to why a Microsoft shop would use Oracle - well firstly Azure does run Linux rather well - But the real reason is for legacy products that would cost too much to migrate to SQL Server and for RAC server - for which SQL failover clustering is not an adequate equivalent for some real mission critical requirements....and of course recent versions of Windows Server tend to scale better than Linux for very high IO requirements.
An awful lot of companies run their payrolls on ADP. It used to be (and may still be, I don't know) that the machine in your server room was a Windows server with an Oracle database, which held the stuff your accounting department worked with, and periodically synced with the big ADP machine off somewhere.
Amazon is certainly strong(ish), but you've nailed the key issue with your comment: Corporate.
It's rather counterintuitive, but you can't actually compare the technical aspects of any two or more products, in any industry and arrive at a useful forecast for market acceptance. There isn't an industry in existence where 'best in class' is actually the market leader. Not a single industry. If there were, it sure as shit wouldn't be IT anyway (sent from my overpriced iPhone :)
As an industry, IT is a bigger slave to fashion than the fashion industry, and those fashions morph into incredibly odd abstractions of reasoned thought in direct relation to the financial size of the client. That's why Enterprise Salesdroids regularly make a lot more money than the CEO's of the companies they work for. They're the only people fucked up enough to think in those utterly senseless abstractions big companies thrive on.
Hell, I fell into IT by accident because I saw somebody flashing a 'SWOR' grid around and I thought I was hilarious. Three months later my buddy and I collected very large checks and new jobs with the company that bought our 'Asset Analysis' package. Haha. Dumbasses. We both still have our original checks, now marked Deposited, in our offices and were both sued later when the company hit money problems. That lasted about 30 minutes in court, but we hadn't done anything wrong, they were the ones who bought the stupid thing; SWOR Analysis being the next 'big thing' that year.
IBM deliberately overpriced everything they sold simply to take advantage of 'nobody ever got fired for buying IBM' bullshit. The very same type of people who thought SWOR Analysis was a good idea, who always bought IBM, carried Palm Pilots and who made Blackberry a successful company, instead of aggregate fruit masquerading as a berry (Watermelons are berries masquerading as melons, really) and it's the same people who fuck with us by naming things incorrectly that will justify their decision to buy a 'Corporate' product that will make MS and Oracle a lot of money. Don't go thinking even large businesses don't like fashion.
When I meet an arse, my best strategy is to introduce him to another such like person. Often as not, they'll take each other out, without any further effort on my part. I don't have to kill one and go to jail for it that way, they do it for me.
Not a big believer in the cloud anyway, as noted here in another article, it's not good for many people.
Not a fan of anything that ignores standards - like Oracle SQL, MS's browsers and so on. Let them "embrace, extend, extinguish" with inferior performing products.
Don't go on about how some slick proprietary add-on to a DB langauge makes it faster. It's not faster than real code, and puts business logic where it doesn't belong in a good design. Good for Oracle DB consultants and support, but not the customer, who now has to buy more hardware to get the same job done.
Paired with a company that thought embedding audio and video into Word (how do I print that, again) and almost security-free remote code execution (OLE, COM and so on), while mixing data with code so that it's near impossible to make secure?
Yeah, that's where my IT money is going - not.
Silly person. If you couldn't embed video in a Word document how do you think the paperless office could be a reality without that feature?
I used to have this really cool wall mounted whiteboard calendar (kind of like those racks where stores display posters) where I wrote important dates for my department. At some point in the mid 1990's our company hired Xerox to 'help us transition' to the paperless office that has proven so successful. After 14 months and several million dollars, I had the calendar moved to the storage room and placed behind all the boxes of blank paper we still had to buy. My request for another CAD person had been declined and I wanted to prove a point. It was a weekly routine to move the boxes to add new info, then put them all back. It was still that way when I left :)
I'm not sure about the whole cloud thing yet. If data security wasn't such a big issue for us I would chuck everything (except the big system we run simulations on) in our data center into the dumpster and increase the size of the staff lounge. I just hate everything about datacenters if they don't belong to someone else. But that's just a pet peeve of mine. On the business side the 'cloud' has a lot to offer.
Cloud is just fantastic for startups. I'm part of a decent size investment group and one of the common factors in every single one of our failed ventures is the cost of tech budget overruns in replacing incorrect tech purchases.
That's not a function of 'bad planning' on their part, it is a function of startups. If anybody in a startup tells you THIS is exactly what we'll be needing in 18 months they're either full of shit, or deluded and doomed. Most startups don't end up actually selling what they thought they would to the market they thought they would. That's just the reality of business.
It is nearly guaranteed that when their CEO comes back and needs a couple million more dollars it's to replace things they didn't need with the things they do need. Those 'incorrect' purchases are quite nearly worthless and the additional funding decision is based, in large part on if I can find someone to give the incorrect stuff to. Schools and NPO's mostly.
But the cloud offers quite a lot of flexibility, and if the company flies they can start to refine their tech needs based on the reality of the situation. Even with the increased costs, that flexibility is still just so very much cheaper than replacing brand new kit with more brand new kit. Cash on hand is the biggest advantage a startup can have. They won't even see 'real' business challenges for 18-24 months at least and if they don't have the money to get through them that's the moment all the dreams of a startup die.
We'll help them find another investor, if that's what they want, but when real challenges appear and they're already out of money, selling us more equity isn't worth anyone's time, or money. If no investor appears I'll generally just wind it all down and sell any assets and IP they've got and cut the losses. Cloud services provide them with a bit more breathing room, and that's huge. We've had projects that went from near useless to making their management team wealthy in just 90 days. There's a shitload of luck required to make a startup successful and every last day you can squeeze out without shutting the doors is one more day to get lucky.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
