Another huge heaving grunt of a move as Microsoft shifts its great fat rear to sit more comfortably on the changing furniture of the computing industry and offer cloud computing services. It could be very bad news indeed for the storage industry. Microsoft's Azure Services Platform - ASP as in 'we've been here before, surely' - …
The Cloud or a puff of smoke?
Does anyone really think seriously that I will trust *MY* data to be stored remotely by the likes of *MICROSOFT*? Ha! If they really think so they must be madder than all those who were buying CDOs for subprime mortgages as if they were the new hot cakes.
Voluntarily sending one's daughter to be a hostage to Taliban is the only crazier thing I can think of at the moment...
The proprietary nature of data
Would businesses want to risk handing over their sensitive data to the cloud?
Never mind the data protection implications.
I think this is why SaaS has never really taken off big time, and it should be the same for storage too.
Bit of a simple approach, not?
OK, so on the one hand you get more users sharing disk capacity on the server (most desktop systems have less than 25% disk usage), but on the other hand the datacenters do and will store the data on RAID systems (RAID5 or RAID10) so those disks will need to be bought too.
And they buy the more expensive RAID certified disks.
Granted, they will pay less for the disks given the quantities, but still...
In other words ...
"We need to lay the groundwork with new types of system and storage applications for a world of parallel computing."
In other words, there are plenty of standards out there but we will build and use our own for the sake of interoperability (and obtaining license fees).
That is how I read it otherwise why aren't other companies, like Amazon, laying the groundwork?
possibly... but myth
"business continuity and disaster recovery software to shift to backup data centres they no longer have, backup software, backup reporting software, security software for files they no longer need to protect, replication software for the data they no longer need to replicate"....
This makes the very seriously flawed assumption that 'clouds' will do this, which we certainly have not seen so far.... put your stuff in the clouds and that is where your risk is in the clouds where companies have only an SLA to flap in the face of a major data problem.
Lets look at out sourcing, and how that was going to remove all the problems of development...
The problems don't go away, they just move somewhere else, often somewhere on the end of a very fancy contract saying the vendor is lacking in any responsibility.... particularly given MS's record in this area!!!
My storage is definitely staying in a place i can touch the disks if i need to.
Clouds are for the marketing dept and board to wee themselves over. The rest of us are still very skeptical.
Not the whole truth
While it is true that moving to cloud computing can reduce the needs for some types of hardware and software, it doesn't totally eliminate them.
Cloud processing can be very good for bursty usage, but doesn't work so well for constant usage, where traditional web-hosting or servers may be better.
As far as storage is concerned, storing using something like Amazon S3 doesn't remove the need for backups - data loss can still happen, either due to problems with your supplier (yes they have SLAs, but the data loss can be more costly than the SLA credit) or accidental deletion/changes/corruption.
All well and good...
but these lot all seem to think that a connection to the cloud is unbreakable and never fails. Having you data stored where you have no means to access it due to a storm or other issue just means you become utterly destroyed due to it. If "somehow" the data centres holding all this data has some major issues (*cough* not like we've seen power stations get flooded in the last couple of years or anything have we), then it's a large number of customers who get screwed over. While keeping the data local gives you a much more scattered approach, and worst off, you maybe the only company down.
Used along side existing systems, as another backup maybe. But used as the sole replacement for current systems it's daft, or security, for reliability, for controlling your own data. It's just a bad idea all round be it from MS, or Google, or anyone else.
"Does anyone really think seriously that I will trust *MY* data........... "
It doesn't matter what you think, nor what the techies say, nor the users..... the thing that will push the Cloud is cost, nothing else.
If its cheaper, and it will be, your business will switch to it.
That is all.
"Mr Ozzie's words were taken out of context"
"We need to lay the groundwork with new types of system and storage applications for a world of parallel computing."
"We need to set the agenda so we can persuade all the schmucks to give us their data and make them pay us for access".
Fixed that for ya.
Setting themselves up for failure.
I read with interest all the changes relating to cloud computing and the massive data centers springing up around the country. While in the Air Force they had a pretty big shift from only a few central main frames to a more widely distributed network to improve survivability in the event of a natural disaster, system failure or hostile action. With the current shift in industry and business to rely on cloud computing centers for the bulk or their computing and storage needs it seems like they may be setting themselves up for failure without maintaining a separate in-house capability. It seems like the concept of a more survivable IT architecture may be a thing of the past.
We already know that occasionally the backbone infrastructure of the internet fails which would leave these companies cut off from their core information. We also know that in this current crazy world that there are people willing to die to cause harm to the USA as evidenced by 9/11. So by relying on data centers and cloud computing to carry the load for business we are making these data centers a valuable target for those with the will and desire to cause us harm. I would hate to be the IT Director for a major corporation trying to explain away the loss of the whole shebang to the board and investors if such a disaster were to occur. Cloud computing and data centers may be the wave of the future but unless very carefully managed they could spell a major failure down the road for people relying on them.
How solid is a cloud?
Agree with most of what the others have said.
Would anyone seriously want to step off a plane onto a cloud (no matter how sold looking it is)?
It seems to me that cloud computing is more about control and world domination than a really serious, 100% rock solid business service.
I have 0% trust in any of the providers - so when I use any of the so-called cloud it's for totally non-critical, frivolous stuff !
Never happening here
Small finance company. We can do some stuff if we lose connectivity for a while, as we have our files, but trusting the cloud? Madness. Utter utter madness.
So the boss will probably recommend it for FY2010
isn't the cloud where lando lived?
anyway, why would i bother when i can buy a pea sized, 200 gazillion gigabyte hard drive for 37 pence? so i keep MY data in MY pocket, not at lando's house.
What about infrastructure costs
This all sounds good until you work through a potential implementation. Infrastructure costs - specifically bandwidth- makes the cloud computing model unworkable for mid-sized to enterprise-sized companies. When you factor that in, it would actually cost you more to outsource your server and storage infrastructure than it would to keep it in house. Companies that run global virtual desktops to a centralized location (their own data center) can easily reap the benefits of centralized management, control bandwidth costs and manage their own security much better than any cloud computing model being offered yet.
Who in their right mind would trust critical data to Microdull?
To boldly go ...
I sense lots of BOFHs worried about their jobs.
Whether your data is stored somewhere on the Internet or not, most businesses are already stuffed if they lose their connectivity. Even if your data is local, if your Internet connectivity fails you can't take or place orders, can't send or receive emails, and increasingly can't even get on your (VoIP) phone. Most business is already dependent on the cloud (or whatever you want to call it) whether they like it or not.
SaaS is already happening as will cloud storage. SLA, reliability, security, and regulatory compliance issues are either being dealt with or will be dealt with over the next year or two. There will be enterprise class storage service providers, probably delivered by IBM or EMC or Sun, as well as the consumer services delivered by the likes of Amazon and Microsoft.
If you don't want to be dependent on a single storage provider, then use several. That's the beauty of it.
In five years, any business who is not using SaaS and cloud storage as part of its IT mix will be seriously uncompetitive. The cost advantages are just way to compelling.
I hate fads and hype. I hated dot com and I hate Web 2.0 BUT cloud computing, although I don't like the name, will transform the way IT is delivered. It's not new technology, it's just a new way of delivering old things.
I bet that when factory owners told their in house technicians that they were going to scrap their local generators and buy electricity from the national grid the techs told them they were mad. Cries of "But what if the power station fails, or the power cables are damaged, or the local substation blows up" must have been heard up and down the country.
It sure is funny how so many of the big guys are throwing the year 2012 and so much blue out there these days with their newer plans for our collective future...
Re: The problem...
"...the thing that will push the Cloud is cost, nothing else.
If its cheaper, and it will be, your business will switch to it."
I'm not rich enough to buy cheap things... The cost will not be the deciding issue.
Eggs -> Basket
You can't shoot a cloud...
As james said... there is nothing solid. Bullets would pass through harmlessly. Cloud computing is like that, there is no central cluster somewhere that can be brought down by disasters.
Sure there are the problems of programming, but, hey, it's microsoft, they kinda know how to write code. I would expect cloud storage to be actually accomplished in a RAID 0+6 sort of way, with the data redundantly spread across geographically diverse locations.
The downside at this early stage is securing the connection to the cloud. A jerk with a baseball bat and a ladder can take out a couple dozen strands of fiber just by whacking it on the outside. (this has already happened to my fiber connection, no visible damage on the outside, many cracked fibers on the inside.)
Of course multiple ring configurations, etc., etc. will keep the enterprise connected.
Nevertheless, the switch to HD and SUPER-HD digital video will require terabytes of local storage, one hour of well compressed 1920 can consume 5-20 Gigabytes, and newer high end production cameras like the RED camera (4K horizontal pixels) generate 60 megabytes per second. A large production facility might have 20 cameras operating. Far too much data to transmit over the network.
Hard drive arrays are the new videotape and the new film. As the enterprise data platform moves to the cloud, the video production marketplace will grow as a customer for hard drive capacity.
Paris because she already populates more hard drives around the world than any other millionairess i know..
Well can't speak for industry as it does really strange things at times (such as adopting Windows 3.1 and earlier) but speaking personally, I am not going to rush to have my data processed off of my own computer on software I don't own to the extent it isn't even running on my machine, but "out there somewhere", in the void.
Who comes up with these hare brained suggestions ? And who pushes them forward as if they are the best thing since unsliced bread ? Bound to be a huge success then.
I think all that £1.8 trillion (and counting) lost by the World's financial institutions lived in the monetary equivalent of a cloud. The whole thing proved to be incredibly vulnerable to a "common mode" failure. That is there weren't lots of single, independent, resilient financial systems. There was one massive one connected by means that nobody understood.
The whole "cloud computing" and SOA idea has the same uncomfortable feeling. Which little common flaw, what single component might fail that trashes the whole thing. There's going to be a complx system of interdepencies, on security, on networks, on naming, on service, on versioning. What failure, benign or malicious, technical or commercial could bring the whole thing crashing down. What undetected security breach will mean that all our details end up in some gangster's hands. I don't think I want my personal or financial data held in this piece of doubly figuratively "vapourware", for what else is a cloud made of. The finance industry made a mess out of opaque and abstract services. This could be a way for the IT industry to go the same way.
So yes - for nice for consumer toys, for the little software luxuries and gadgets of computing, for stuff not emeshed in the working of the real economy than mash these things together. Just don't bet your life, or your finances on it. If you want guaranteed security, performance, robustness, accountability and reliability. Well that's a lot of trust to have - a systemic failure could bring down a country, not just a company.
Been here before...Haven't we.
Time was, in those far off halcyon days of the 1970s, when companies bought time on mainframes run by Bureaux, Centrefile springs to mind.
This was because computers were expensive, then along came mini-computers which reduced the cost of ownership and operation, and got rid of all those problems associated with remote computing. Like the mainframe not being available, leased lines out of action, running out of capacity, loosing your data and so on.
Can't see the cloud is any different, just far, far bigger and more complex, with far more issues.
Sadly the decision to go with the cloud will be CEOs & CFOs who have been lead by the suppliers who will sell to them and not their technical people. It will be sold as being cheaper than doing it yourself, and as a way of removing a load of management issues so that the management can concentrate on their core business activities.
Company board members are stupid, ill informed and naive about much of business, not as individuals, but as a breed, they have little time or incentive to be much else, other than keeping an eye on how to make this years bonus.
Clouds are outsourcing in sheeps clothing
This is just the next flavour of outsourcing isn't it?
Really though how many large companies will trust MS or many other vendors to run their remote DR etc if fthey've got stringent data retention or security policies in place?
I can't forsee any company who has gone through an outsourcing exercise, then finding out what a **** up it is, insources once again would ever go for cloud computing.
There are companies now who when they have a drive fail in an array don't let the vendor take the failed drive away when it's replaced, but rather spend the $$$ destroying it.
Until I see IBM and EMC wading in, I doubt many are going to look at it. Outsourcing already is cloud computing, no doubt the outsourcer trying to run the solution in the most efficient (read cheapest) means possible. Yes there's virtualisation, yes there's de-dup, but with the holistic policies of many companies regarding data retention and paranoia large shops will just never go for it.
I also find it quite ironic that we have one article on here sounding the death bell of storage management and another with a whitepaper harping on about the growth of storage management. It's like the arguement about the mainframe platform being a dinosaur. Admittedly it's quite complex and outdated in some ways, but it is extremely efficient at what it does and does it with a robustness that many platforms fail to even glimpse at. Storage will not eventually go to the cloud and to just a handful of vendors because the cockups that could be made would be of a monumental proportion leaving many companies to pick up the pieces of what they thought was data being managed with a high level of integrity.
Buy cheap, pay twice.
its the plans for Windows that concern me
As I'm not paying out for bandwidth to support hosted Windows
It's all about Trust
The first thing that needs to happen is 'Trust'.
I guess we're at a similar stage hundreds of years ago when we didn't trust Banks. (OK, we still don't trust them, but we have to use them!). Trust needs to be earned and these early developers better make sure they get it right if they want anyone to adopt this.
However, I'm surprised there is no mention of the two biggest potential UK Storage providers - BT and Virgin Media. For Cloud to work it needs a reliable fast pipe and only these two companies can supply it. Why wouldn't they want to set up their own services? They can offer two things that serious Gamers have known for years is supremely important - pipe speed and low latency.
Then there is the potential to supply Joe Public with the Cloud.
I bet a simply box with Video On Demand followed by a remote OS with online storage for all of Mr Public's photos and documents would be an easy sell to him.
Imagine no more vaporising of his data because he's zapped his hard drive by doing something stupid!
Yes I'm talking about the NC concept that failed to lift off due to the lack of a decent pipe. Well we've got the pipes...
If someone can make Cloud work for Mr/Mrs Public and family, then that'll develop the trust needed for businesses to seriously take the plunge rather than just tinkering with this new service.
So I'd keep an eye on companies like Virgin Media. They've been supplying Cloud to their customers for years in the form of video on demand and some of us don't even recognise this as Cloud, even though it is...
Quite why Virgin Media don't offer a netbook with 3G/Wi-Fi box already is beyond me. It'll be a gold mine.
Microsoft are pretty screwed I think
Yeah yeah the cloud this, the cloud that, but M$ think are pretty screwed aren't they?. No one likes them or trusts them and they churn out complete poop like fi$ta? They jump on bandwagons far too late and make a horrible product (or buy a good product then make it rubbish)? Can you imagine - a M$ cloud? Hotmail is barely on par. Still, I suppose they are a few levels ahead of Yahoo! in terms of how long it will take them to disappear.
The "cloud" means MORE storage bought, not less. Storage in the cloud PLUS local back-up (as the punters foolish enough to fully rely on cloud storage are not alot).
Sick it in the cloud? Not on my watch...
Sorry, I don't agree that cloud computing is the Next Big Thing (tm) or even the Current Big Thing.
The main problem is one of reliability of systems not under your direct control and the associated risks to the business. A company running its own servers and infrastructure is in control when and if an outage occurs. As such it can employ whatever work-arounds and solutions it wants when something goes down.
If a company moves its servers into the cloud there are now three separate infrastructures - the remote data centre, the local company infrastructure and the network linking the two. The company has ceded control of two parts of its overall infrastructure and has no direct control over it. If something goes wrong all the IT support guys can do is... call another IT support team and wait.
I'm simply not convinced by either Amazon or Google that they can achieve the uptime I'd get by keeping everything local, with the same degree of control over the whole infrastructure, particularly in the event of a failure of some sort. And I certainly don't anticipate Microsoft performing as well, if only because their cloud is based on Windows technology.
It will only take one significant outage for companies that have not undertaken a risk assessment to realise that they have no control over a business critical IT system and for them to change their minds.
Will people use cloud computing? Yes, if the system is not business critical. Yes, if the system is business critical but the company is establishing itself and can't afford to purchase run its own hardware. No, not if the absence of a system for anything more than a few minutes stops the business functioning. The risk is to great.
And I've not even thought about data protection, security, etc, etc...
Cloud = Vapour in suspense?
So not content with screwing up the desktop, MS decided to go for the corporate throat big style...
But its not just a matter of trusting with data, as I understand it, its all of the functionality that goes with it. I mean at least with a local server that goes tits up & stuffs all your data at least you generally have something from which you can recover,
Or perhaps Im just too old school to accept that theres 'new ways to do familiar tasks', which, lately seems to translates to "it used to be straight forward, so weve fucked it up good and proper"
Perhaps someone at MS could try 'familiar ways to do new tasks'? It might just catch on...
"I bet that when factory owners told their in house technicians that they were going to scrap their local generators and buy electricity from the national grid the techs told them they were mad. Cries of "But what if the power station fails, or the power cables are damaged, or the local substation blows up" must have been heard up and down the country."
This is true to a point; because it looks like the owners heeded those warnings:)
Think about it: Most folks still buy UPS systems, generators, etc. Why is that?
Because the Grid fails, even redundant Grids.
Cloud computing is useful for many things, but it is not a Panacea; it's a main frame dressed up in new clothing. The power of cloud computing doesn't mean that you'll be outsourcing computing needs ; on the contrary it simply means that savvy shops can build the same main frame type environment on the cheap.
Many places many devices model
The cloud is not about taking stuff off your personal computer or internal network, its about having your data and applications anywhere, irrespective of device or where you are.
If you have a single or even just two computers then the cloud doesn't make too much sense. But once you realise that some people work off 5-6 or more computers in different locations (saves carrying them about and the risk of loss or damage) and then need access via non-computer resources such as mobile phones, or want to connect devices like cameras directly, and want to be able to share and distribute that data to others then the cloud is very strongly the direction things are going. You can't do it non-cloud - what are you going to do, walk about with a memory stick and hope you don't lose it and that it doesn't get stolen?
Yes, big corps with big data centres already have some of this type of stuff running internally under remote access. But it costs. If people like M$ (and I don't think M$ will be the only people offering these M$ services because of competition rules and issues like some people don't trust M$) run these services then the cost will be strongly driven down.
And the data will be distributed/mirrored with redundant backup and sync'd to your offline hard-disk. M$ is betting very very very big on this model of computing. These small-scale announcements and press releases really aren't doing justice to the scale of the Microsoft gamble.
I don't think so
customers drawn to cloud computing would never buy an EMC, HDS or Netapp array in the first place. they're more likely to have their data sitting on direct attached storage running RAID0 on 1TB SATA disks, and have 1 poor schmuck doing everything. So what difference does it make if they dump their crap into the cloud. But customers with serious apps and mission critical data surely wouldnt move it to the cloud. Also, even if all the companies in the world decide to entrust their data centre operations to Microsoft (god forbid), their data has to sit on some form of storage. Whichever storage vendor does business with Microsoft is likely to benefit. So not ALL storage vendors will lose out. Some likely to gain from this.
Maybe Sun will put Fishworks in the Cloud! Bring ZFS TO THE WORLD!!!!
I am beginning to think.... that just like everything else on this planet..... that the IT Industry is cyclic. Back in the mainframe days - you could call that a cloud. Then there were the data processing bureau's then a huge devolvement (PC's and Intel x86 servers) now we are heading centralised again albeit in cloud form.
I think it will be a success - more than likely because of marketing and price..... BUT there will always be companies that want their own infrastructure to look after their own data and run their own apps on rather than trusting a third party cloud that is 'shared'.
"It doesn't matter what you think, nor what the techies say, nor the users..... the thing that will push the Cloud is cost, nothing else. If its cheaper, and it will be, your business will switch to it."
Utter nonsense. At best cloud computing is an alternative backup regime. Data is the lifeblood of an organisation. Amazon, Google, MSFT etc are not stupid. Cloud access will be cheap to encourage adoption... but once you are locked in, don't be surprised when the subscription price rises. And what happens to your data if you stop paying? Adios!
You need to be in control of your own data. Just watch the cloud bubble burst with the first reported data loss.
Sorry folks, but the "someone cut through our fibre optic cable in Docklands, the Internet's down, all servers are offline because of it." scenario kinds screams out at me..
Gimme a good ol' fashioned gigabit network connection anyday..
Re: The problem
> If its cheaper, and it will be, your business will switch to it.
> That is all.
Risk is a cost, downtime is a cost.
Maybe your business will forget this and switch - if so they are managed with the same ethos that will also dump the service on the first significant c*ock up.