Microsoft Azure Outage Across the Globe 167
hawkinspeter writes: The BBC reports that overnight an outage of Microsoft's Azure cloud computing platform took down many third-party sites that rely on it, in addition to disrupting Microsoft's own products. Office 365 and Xbox Live services were affected.
This happened at a particularly inopportune time, as Microsoft has recently been pushing its Azure services in an effort to catch up with other providers such as Amazon, IBM, and Google. Just a couple of hours previously, Microsoft had screened an Azure advert in the UK during the Scotland v. England soccer match." (Most services are back online. As of this writing, Application Insights is still struggling, and Europe is having problems with hosted VMs.)
This happened at a particularly inopportune time, as Microsoft has recently been pushing its Azure services in an effort to catch up with other providers such as Amazon, IBM, and Google. Just a couple of hours previously, Microsoft had screened an Azure advert in the UK during the Scotland v. England soccer match." (Most services are back online. As of this writing, Application Insights is still struggling, and Europe is having problems with hosted VMs.)
Out of band patch.. (Score:1)
Really makes me keen to install that out-of-band patch that was spoken about yesterday across all my servers....
Re: (Score:2)
Re:Out of band patch.. (Score:5, Informative)
I installed it last night on all domain controllers after testing it in my isolated testing network. It's not really optional since it allows any domain user to become domain admin and the only resolution to that is a domain rebuild or authoritative restore. It's also already been seen in attacks in the wild so you can assume the next client to get driveby malware will be going for domain admin.
Re: (Score:2)
Re: Out of band patch.. (Score:4, Insightful)
Regardless of OS, poor testing of third party apps / services or poor security as part of your deployment, can cause you to be violated. I have seen many Linux server still using Telnet or VNC for management, and allowing ROOT to login directly to them....
Secure your environment regardless of what you run......
Re: (Score:2)
We'll hear, next week, about a zero-day that takes down Azure. Oh, wait.....
Azure is blue, ain't it? (Score:3, Funny)
Global BSOD!
Re: (Score:2)
Global BSOD!
Happy little Blue Cloud of Death
Yawn ... (Score:5, Insightful)
Cloud fail, like nobody saw that coming.
If you don't own and operate your own infrastructure, you're at the mercy of someone else.
And clearly that someone else can't guarantee you robustness with this magic cloud.
All of these people who say "awesome, because, cloud" -- well, I have yet to be convinced that any of these vendors can provide as much uptime and reliability as a decent IT department.
I suggest we start calling it Clown Computing -- you cram a lot of Clowns into a tiny little car, and hope it keeps going.
When something goes wrong, hilarity ensues.
Re:Yawn ... (Score:5, Interesting)
Yeah, but it's never really been about the reliability. It's always been the "not paying your own IT maintenance staff" thing that's the big draw.
Re: (Score:3)
Sure, but when you have outages and stability issues which impact your business, is it really a good trade off?
I mostly see this as a management fail -- penny wise and pound foolish.
I will be curious to see what percent of companies who went to the cloud will transition back to doing stuff in-house, and just how much that will really cost them in the long run.
Re: (Score:2, Insightful)
Right. Because in-house infrastructure never fails.
Power outages never happen.
Lines are never cut.
Patches never fail and rollbacks always work. ... can I come live in make-believe-land with you?
Re: (Score:3)
Re: (Score:1)
When we talked to an MS rep about the physical infrastructure at Azure, they told me they built in scale with commodity hardware to keep costs down. This means servers with single power supplies, etc., none of the common redundancy stuff you would generally use in a real install. So, to address that, you need to pay them double to have two VM's up to keep your services up.
Re:Yawn ... (Score:5, Interesting)
There is something you can do about all of those conditions.
With cloud, you just wait for the rain (outage). You can pray (call an outsourced tech support department) for it to stop raining (services restored), but until god (cloud provider) decides the rain is done (fixes the problem), you're getting wet (offline).
That gives me a new "Cloud" tagline:
Cloud - We will definitely rain on your parade.
Re: (Score:2)
Re: (Score:1)
I'd prefer the failure to be inhouse any day. If something fails on premises, we don't have to rely on someone else to tell us what went wrong. We don't have to operate on someone else's ETA for service restoration, we have eyes on the process and know how long it will take. We don't have to hope and pray that a third party fixes whatever problem they have internally so it doesn't happen again, we address it ourselves.
More to the point, if a failure happens because of my fuckup or my team's fuckup, I can qu
Re: (Score:2)
Right. Because in-house infrastructure never fails.
Power outages never happen.
Lines are never cut...
If the power goes out in your company building, odds are perfect that your users are going to be sitting in front of dark workstations long before the UPS gives out and the servers shut down. ;) Same with most general IT outage situations... if a patch borks your 'doze servers, it's likely going to bork your 'doze workstations. If your Internet/WAN is dead, it's going to affect your users too. Unless your users are all remote and on VPN, a local problem is going to affect your users in more ways than just t
Re:Yawn ... (Score:5, Insightful)
However, in a widespread outage like this, I'll bet the big cloud providers have a better record of rapid recovery than their customers had in-house. By necessity, MS, Amazon et al have very competent engineers who know the product well available to pull off what they are doing (including sleeping) and jump into any really serious problem. There simply are not enough such engineers to go around all the mid-sized IT organizations in the world nor interesting enough work to keep these engineers interested and sharp at most of these IT organizations (to say nothing of the cost of keeping such engineers around).
For a car analogy... When your high end car has a nagging problem that your local mechanic can't figure out, the dealer often can figure it out quickly, possibly with the help of a factory specialist who deals with (say) ECUs on only this make all day, every day. Rarely can an independent mechanic specialize enough to come close to the factory specialists in diagnosis. Now, if your car just has a dead battery, your local mechanic may give you faster, better, and cheaper service than the dealer.
Awesome! wait on experts so we can run again (Score:4, Insightful)
Are you kidding or do you not understand how large companies, in particular cloud companies, operate? Have you ever had to call one about an unknown issue? Try it sometime....you'll learn a lot.
Re: (Score:2)
Large cloud companies treat large clients wonderfully. Large cloud companies treat midsized clients like crap. Midsized cloud companies treat midsized client quite well. This is the sort of thing you should be discussing with your cloud agent, getting you into a rightsized relationship with your vendor. Which BTW is also likely to save you money.
Re: (Score:3, Informative)
Re:Yawn ... (Score:5, Insightful)
The calculations are simple when you assume the cloud will fail and your infrastructure will not. A real tradeoff calculation has to include estimates of the reliability of both scenarios. The answer to "Is it really a good trade off?" will be entirely based on estimates and opinions. I'm not saying you're wrong, I'm just saying that the math does not spit out "no-brainer".
Some cloud providers will even give you SLAs with real money behind them. So, they could conceivably come up with a no-brainer deal where the cloud provider guarantees your $80,000 every day, whether it's from having your business up and running or writing you a check.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
--The "cloud" is no substitute for decent IT admin. Even with hypothetical 100% cloud usage, the company should be keeping *at least* 2 decent admins on staff for planning (resource usage, disaster recovery, rollouts, etc), which-holiday-are-you-working swaps, and vacation coverage (not to mention sick time and accidental death-type stuff) as well as you know, daily work...
--And that's the bare minimum! Any less than that and the poor soul will burn out and eventually quit, or get sick.
Re: (Score:2, Insightful)
I spent 6 months earlier this year on behalf of our IT Director (who wanted us to go to cloud really badly, because, well, cloud) studying the costs and efforts of doing so. My conclusion was that over a 5 year period, cloud hosting would cost us TEN TIMES the cost of hosting internally. I expected this report to end this discussion, but it didn't.
My director pointed out I hadn't taken into account the fewer people we would need to manage things (which I pointed out was horseshit, we do colo hosting now a
Re: (Score:2)
it comes down to the accounting lure of a low monthly fee (operational costs) vs high one-time costs when you buy equipment (generally capital costs)
rent the servers then. Most server manufacturers will rent them either directly or on a 3-year rent-until-obsolete-and-replace contracts. It costs more over time (especially if you keep them running for decades) but it does what you want for the beancounters.
I think people want to go cloud because of all the hype and advertising. Your boss will have read lots
Re: (Score:2)
FWIW it is fairly easy to create a structure to turn internal capex into opex. You can still do stuff internally, have your people working being paid via. a PEO / staff aug firm and have the hardware leased back to you. If the issue is they want opex that's easy enough to achieve under either model.
Re:Yawn ... (Score:5, Insightful)
Sure, but when you have outages and stability issues which impact your business, is it really a good trade off?
Of course it is. Outsource to the cloud and cut the quarterly costs massively by laying off staff. Get a big bonus. Possibly share options go up due to better profits and blathering to the shareholders about the cloud. Sure 3 years down the line it might tank for a few days and in one fell swoop wipe out all the savings and then some.
Not my problem, I'll be long gone.
So is it worth it? Hell yes!
Re: (Score:2)
Not my problem, I'll be long gone.
...and at another company which did the same thing, except this time it's your head on the chopping block (even if it isn't your fault.)
Saw that happen to an IT director once. It was messy, to say the least...
Re: (Score:2)
It depends. Sometimes, yes.
Re: (Score:2, Insightful)
This is why I drill for my own natural gas. You never know when the gas company will have an outage.
Re: (Score:2)
There's ways to mitigate it. I'm not familiar with Azure, but I do work in Amazon EC2 all day - we spread our virtual private cloud across 3 availability zones in one datacenter, and use a VPN tunnel connected to another virtual private cloud in another datacenter on the other side of the continent. Because everything for standing up instances is automated, we can rebuild our whole platform in about 30 minutes including data and applications on the other side of the country.
Yes, there is still points of f
Re: (Score:2)
Sure. You get to blame your cloud provider for their problems, and with luck for even some of your own. You get bonuses for the savings and a scapegoat for the costs. What's there to trade off?
Re: (Score:2)
I agree that many have simply jumped onto the cloud bandwagon because it's new. Some cloud providers off
Re: (Score:3)
Yeah, but it's never really been about the reliability. It's always been the "not paying your own IT maintenance staff" thing that's the big draw.
I priced 10 2core VMs. It was 24k/annum. We do that internally on an R720 that cost 10k and needs about 3 hours a month maintenance. So for mainly internal use networks, where is the value?
Re: (Score:2)
But, but, ... it's the CLOUD.
Off course you will spend more, you have to account for the overhead (sales, marketing, support) of the hosting companies and their profit.
Re: (Score:2)
I priced 10 2core VMs. It was 24k/annum. We do that internally on an R720 that cost 10k and needs about 3 hours a month maintenance.
Sounds quite expensive: that's only 20 cores. You can get a 64 core 1U commodity machine from super micro for under 10K, with a decent chunk of RAM too.
Re:Yawn ... (Score:5, Insightful)
Sure, cloud computing has its short-comings. But it has also allowed a litany of small companies who simply can't afford to own their own infrastructure to do business.
Re: (Score:2)
Excactly, that is the key here. It does have its purpose and all too often these discussions focus on the 'Cloud vs Inhouse'. It's not that simple. You have weighed the pros and cons and have determined that cloud computing is more advantageous for you and probably a whole lot more cost effective. But its not a fit for everyone and I suspect cloud services over promise to some clients for which it is not a good fit.
Re:Yawn ... (Score:4, Insightful)
Once again, missing the point. In my (small) shop, by using azure (which has worked well for us), we avoid having to use money to hire admins to maintain any sort of in house servers we might have.
Who maintains your Azure infrastructure (I hope you built in all that lovely redundancy for these problems) and how often do you really need to maintain internal servers? If these are on 24x7 you're going to be paying through the nose and if you miss a monthly fee, off you go. Not to mention that cloud servers are horrifically under resourced compared to hardware you can buy, so you generally need many more of them, and none of the bandwidth, I/O or CPU resources are guaranteed to be yours no matter what your meaningless agreement says.
We can then put that money towards more developers (or better salaries for us current devs), as well as paying for training, nicer dev machines, etc.
Ahhh, yes. Developers who believe deployment can be bypassed as a cost and running applications in production (which is kind of important to any company running web applications and who relies on them for income) simply doesn't matter.
At the same time, if we do have a problem with any sort of hosted service through azure, support is literally a phone call away, and I can't remember the last time a resolution didn't happen within a couple hours.
You've been exceptionally lucky, or you're being economical with the truth ;-).
Sure, cloud computing has its short-comings. But it has also allowed a litany of small companies who simply can't afford to own their own infrastructure to do business.
I've also seen a litany of small companies go out of business with cashflow issues who thought like that. Funny that. Yes, the infrastructure is cheaper if you don't run it all the time. I think I once calculated that if you have a server on for more than eight hours a day then you're simply being milked for a monthly fee.
Re: (Score:2)
If these are on 24x7 you're going to be paying through the nose
Check out he prices for EC2 reserved instances, if you know you'll need that server for 3 years. Prices are similar per core to buying entry-level Dell rackmount servers with 3-year support contracts. Of course, the physical Dell has more memory and disk than the VM with the same core count, so you come out ahead there if you needs lots of memory, or local disk, but not by a lot.
Re: (Score:2)
Check out he prices for EC2 reserved instances, if you know you'll need that server for 3 years.
If I was committing myself to a server for three years then I'd buy one where all the resources were guaranteed to be mine...... The whole point of the 'cloud' is to get yourself away from long-term commitments, move around your infrastructure and upgrade as necessary. The fact that Amazon, and others, have started doing this to look better against dedicated hardware tells me that things are not sustainable in that castle up in the sky.
Prices are similar per core to buying entry-level Dell rackmount servers with 3-year support contracts. Of course, the physical Dell has more memory and disk than the VM with the same core count, so you come out ahead there if you needs lots of memory, or local disk, but not by a lot.
I don't think comparing and equating dedicated hardware to a transient A
Re: (Score:2)
The whole point of the cloud is to remove yourself from such long-term commitments
Not so much. The whole point of the cloud is to remove yourself from long-term IT salary commitments, really, and to be able to just write a check for dependable IT quality (which apparently Azure isn't selling this year). Companies that see their IT staff as reliable and inexpensive wonder what all this "cloud" nonsense is about, while companies who just write ever-larger checks to EDS et al and shit still breaks all the time in the data center see the cloud as a wonderful escape.
Re: (Score:3)
We can then put that money towards more developers (or better salaries for us current devs), as well as paying for training, nicer dev machines, etc. At the same time, if we do have a problem with any sort of hosted service through azure, support is literally a phone call away, and I can't remember the last time a resolution didn't happen within a couple hours.
Who's this we? Are you some kind of dev-only shop, self-managed?
I would bet that in most instances, the "savings" from moving to cloud never becomes more budget for the IT department, especially if its money for salaries. If anything it just cuts your budget or feeds some bonus pool for executives.
Re: (Score:3)
Cloud fail, like nobody saw that coming.
If you don't own and operate your own infrastructure, you're at the mercy of someone else.
Pretty much anyone with a brain saw it coming. That doesn't stop a lot of idiots who bought the shit sandwich from feeling burned.
And clearly that someone else can't guarantee you robustness with this magic cloud.
Nope. Because most of the time, unlike when you control your infrastructure, you have exactly ZERO way to verify claims regarding robustness of service.
All of these people who say "awesome, because, cloud" -- well, I have yet to be convinced that any of these vendors can provide as much uptime and reliability as a decent IT department.
And keep waiting. Because they can't. Flat out.
I suggest we start calling it Clown Computing -- you cram a lot of Clowns into a tiny little car, and hope it keeps going.
And my mind immediately flashed to Michelle Duggar.
"Oh! We'll take whatever uptimes God sees fit to grant us!"
When something goes wrong, hilarity ensues.
Unless you're the poor sonofabitch it's happening to. Then it ain't
Re: (Score:2)
I've seen the clown car/duggar meme too. When I showed it to my wife, she hated me for it.
Re: (Score:2)
I've seen the clown car/duggar meme too. When I showed it to my wife, she hated me for it.
You should have immediately asked her if she wanted to be the mother of 14 kids. That SHE had to give birth to.
I'm pretty sure the response would have been "Hell no!"
Re: (Score:2)
All of these people who say "awesome, because, cloud" -- well, I have yet to be convinced that any of these vendors can provide as much uptime and reliability as a decent IT department.
Unfortunately, also from my anecdotal evidence, decent IT departments are few and far between. Budget constraints have gutted them and few are willing to pay the premium for a good admin. They prefer to hire an "entry level" admin and set their salaries accordingly.
That's w
Re: (Score:3)
Re: (Score:2, Troll)
Re: (Score:2)
I have yet to be convinced that any of these vendors can provide as much uptime and reliability as a decent IT department
A good place to start would be getting the numbers. Do you know any numbers about uptime of web-servers maintained by IT departments?
Re:Yawn ... (Score:5, Insightful)
1) You're too small to afford enough full time IT
2) You can't afford the capital investment into your own servers
3) You need a low latency global CDN like service, but you can't afford dedicated servers running everywhere
4) You need only temporarily need to scale up your servers to handle burst load
5) I'm sure there are other reasons.
Re:Yawn ... (Score:4, Interesting)
I don't think anyone is disputing that hosted online services are both useful and, in some cases, absolutely essential, especially for smaller businesses. Well, maybe some people are, but they're pretty much Luddites, so we can ignore them. It's just that in the rush to push everything to the cloud since that's seen as some sort of panacea, people tend to forget that there are serious consequences to outages, and the more you push services to the cloud, the greater the impact of those outages will be. It's essentially putting all your technological eggs in one basket.
As much as people complain about proprietary file formats, those really don't hold a candle to proprietary services as far as vendor lock-in. If the service you chose, for instance, starts to go south on a regular basis, and you've built your entire ecosystem inside a specific vendor's cloud, you could be in a world of hurt.
That being said, my feeling is that these sorts of system-wide outages are simple part of these services growing pains. Even now, keep in mind that these sorts of large-scale failures are rare enough that they make international headlines. In another five to ten years, it's going to be even rarer still. Otherwise, fewer large players will trust them for critical infrastructure over the long haul. For smallish businesses, even with occasional outages, it's still probably a net win.
This is the fatal flaw (Score:2)
This right here is why I don't use cloud services and do everything I can to make them unattractive to the users. The more "investment" made into a given cloud system, the more "pain" received when the cloud goes down. As things currently stand, that means I don't trust the cloud for anything other than basic commodity
Re: (Score:2)
It is also silly to want to get services from only a few large players.
Re:Yawn ... (Score:5, Interesting)
Let me explain it from my point of view. I own and operate a one or two man software company that also hosts web sites. I work in the flim & tv music industry, meaning I have a shit load of music (literally terabytes) that has to be available for download.
8 years ago I owned a rack of servers downtown here that I managed myself. Honestly, it wasn't that bad. I bought reliable used 1U servers (mainly IBM and Dell) off ebay and stocked them with disks. I ran FreeBSD and Linux, used RAID, etc. But I always had two issues to deal with. The main one was "I have to always be available to handle hardware issues".
My company isn't big enough to hire someone to do it, but I managed for nearly 10 years with no disasters. In that time I had a motherboard crap (when I was starting out with one server - ouch) and a few disks fail. In all of those times I had to go in - sometimes in the middle of the night - and fix/replace whatever was wrong.
Then I found Amazon AWS. Here's the kicker - it was actually cheaper for me to simply "rent" storage from them than to rent rack space for my own servers. I moved my servers to linode.com - again it was cheaper although they're nowhere near as fast as my former dedicated servers were, but they're fast enough for my applications and I can always move to larger instances where needed. And that eliminated my maintenance issues for hardware while costing less per month and maintaining the same 3-4 nines level of availability that I've always had. Oh, one other thing - S3 makes it just as easy to secure my audio files but the delivery speed can easily saturate any pipe that the files are being delivered to.
So the cloud might not be "magical" and solve all the world's problems, but for small IT shops it's great. Everything I do is on the internet so the whole "what if your connection goes down?" issue doesn't exist for me. I do not recommend such a solution for everybody. I have clients in the industrial wholesale space and their inventory & sales system definitely should be on-site with off-site backups. But their web site can be hosted elsewhere.
Anyway, yes, the "cloud" is very useful for many businesses.
Re: (Score:2)
Why cloud ?
Why not just a hosting provider for example ?
It's cheaper too:
http://vultr.com/ [vultr.com]
Re: (Score:2)
Linode is a standard VPS provider. I don't need "hosting" where the sites are on a machine with 1000 other web sites. I do high-end business-to-business sites that need to be available and very responsive. Because of using a variety of software pieces and having to run cron jobs and all that I also need access to the machine.
As for disk space there's no comparison between a service like Amazon S3 (or similar offerings from Google, Rack Space, et al) and a bunch of disk space. Disk space is actually pret
Re:Yawn ... (Score:5, Insightful)
Yes, the "cloud" servers sometimes have outages. So do managed hosting providers. So do internal servers. And frankly, although every business thinks that what they're doing is super-important and they can't afford even the briefest outage, the fact is that most businesses can.
If Azure or AWS go down for an hour, it makes news and everyone freaks out because a lot of people are using them. If your business's server goes down for an hour, it does not make news, and people don't freak out. But for the business experiencing that 1 hour of downtime, what difference does it make whether they own the hardware or it's in "the cloud".
Re: (Score:3)
If your business's server goes down for an hour, it does not make news
Not everyone is in that boat. If there is a system outage with the systems I deal with it will make the news, sometimes even the national and international news [wikipedia.org]. That problem wasn't with one of the systems I deal with or was provided by the company I work for but was a wake-up call to the industry.
Re: (Score:3)
What it boils down to is whether the cloud service is more reliable than doing it in-house - which has more downtime? Can you do it better than Azure? The cost then comes into it - can you do it better for less money? The only no-brainer is the service that is both more reliable and cheaper, otherwise you're looking at tradeoffs.
For some small businesses, cloud solutions may be both cheaper and more reliable than doing it in-house, especially if the core business is not IT related.
Of course, that assumes th
Re: (Score:2)
When something goes wrong, hilarity ensues.
sure, because nothing ever goes wrong in the "own everything outright" world. Nobody ever goes on holidays, the right guy is never off sick when you need them most and of course, there's always enough money to make all the right decisions in relation to performance and redundant equipment.
IMHO, whichever way you go, there will be drawbacks. Azure (and Google, AWS, etc.) outages are newsworthy, that's a hint right there. Just keep track of these events carefully so when the time comes you can try to justify
Re: (Score:2)
Owning and operating your own stuff doesn't mean you won't have outages. I have no idea why you would ever think that.
Re: (Score:2)
I really wonder how people make these arguments?
Your in-house servers have never been down?
Every company I've ever worked for has had their in house server go down (email, collaboration, web servers, git...).
Pure anecdotally, but as I've seen cloud services used more, I've experienced less downtime. Yes, Google may go down for a few hours. Yes, Azure may go down for a few hours. Yes, hosted git (BitBucket) may go down.
But these downtime are much less frequent than the internal server. Sure, it might cause s
Re: (Score:2)
Your decent IT department doesn't have dozens of redundant colos attached to your network in class 2-4 datacenters. There are statistical measures of reliability for fortune 1000 companies regarding colo and the colo providers crush their internal. There are statistical measures of reliability regarding commercial cloud and they crush internal. Now that's not saying the best
Re: Yawn ... (Score:2)
Re: (Score:2)
Lots of places, particularly small businesses, have problems with decent IT departments. A decent IT department will include people expert in several different technologies and platforms, enough to cover for each other, and management sufficient to recognize who is good at IT and who is bad. It doesn't scale down well.
Re: (Score:2)
Re: (Score:2)
What?
Re: (Score:2)
Pay him no mind. He apparently left his critical thinking processes on Azure.
Re: (Score:1)
Yet I don't remember the other clown providers making headlines like this. Microsoft seem to have a tradition forming all by themselves.
No need to jump to conclusions just because your memory is poor. A quick search (http://www.lmgtfy.com/?q=cloud+outage) shows several outages across several providers. Microsoft and Amazon are the biggest cloud providers so of course they are going to make the biggest headlines, but certainly Microsoft is not exclusive in this tradition.
Re: (Score:2)
Yet I don't remember the other clown providers making headlines like this. Microsoft seem to have a tradition forming all by themselves.
Then your memory is very short lived. Amazon, Google, Apple, Dropbox, and others have all had very notable cloud outages over the past few years.
Re: (Score:2)
Of your list, only two are cloud providers, the others use other cloud providers. DropBox works using Amazon S3 and AWS. Apple's iCloud works over Azure.
Dropbox or iCloud dying independently of the underlying cloud provider has zero to do with the cloud as they were application level failures. And application level failures will happen regulardless of if it was don
You Still Need Geographic Diversity (Score:4, Interesting)
Re: (Score:2)
Sure, the ones that have to do with infrastructure problems.
But not any of the software originated ones.
Re: (Score:2)
Wow, I'd be pretty angry (Score:5, Interesting)
Everyone forgets that Azure is a way-beyond-massive Hyper-V implementation, and that AWS is a way-beyond-massive Xen-like-thing implementation. Even though both cloud providers let you be smart in designing your infrastructure (multi-site, redundancy, etc,...the tools are there) nothing will save you from an outage of the core guts of the system. Wasn't Azure's last failure due to a certificate expiration? There's no way an end customer can plan around that.
I'm a big fan of the private or hybrid cloud version of this fad. You get all the good stuff that Azure and AWS customers get like dynamic provisioning and software defined networking, without having to rely on a third party. Unfortunately, CIOs and other execs just see the numbers on a spreadsheet and don't take the costs of outages that you can't control into account. Power fails, networks drop, and people do stupid things in on-site implementations also. But you can at least have your staff working on it with the incentive being "you get to keep your job." With a public cloud provider or even a hoster, the responsibility ends with "oops, here's 7 hours of free service" and you have to wait in line with everyone else.
Re: (Score:3)
I'm a big fan of the private or hybrid cloud version of this fad. You get all the good stuff that Azure and AWS customers get like dynamic provisioning and
Not really. I mean sure you get dynamic provisioning right up untill you completely run out of capacity. The advantage of Amazon is that they are much, much, much, MUCH bigger than you. So, if there's a big peak in usage for some reason, you can keep on scaling up to match the demand.
Re: (Score:2)
I wonder how big an organization would have to be to actually make a private cloud useful. As you say, part of the benefit is having access to infrastructure that you don't have to pay for unless you need it. If you own the cloud than you have paid for every last piece of iron that is sucking up power in your datacentre. Does dynamic provisioning offset this cost? Would governments be the only clients that private clouds truly make sense for?
Re: (Score:3)
Would governments be the only clients that private clouds truly make sense for?
Nah, we're on the small side of the S&P 500 and our "private cloud" has enough spare capacity to bring entire new projects online, spin up testing instances, provide an entire parallel Citrix farm (we're upgrading and want to have the old farm available for fallback in case we hit a critical bug), and still provide for the failure of up to two hosts without any overprovisioning. Infrastructure hardware and operating costs are
Re: (Score:2)
Those numbers surprise me. thanks.
Re: (Score:3)
A good devops team means that amazon vs. local is a flip of a command line parameter.
Re: (Score:2)
You're just pulling that statement out of your ass. Most people who run large companies aren't stupid, and I'm sure that many of them do take into consideration the costs of outages.
Re: (Score:3)
"Most people who run large companies aren't stupid, and I'm sure that many of them do take into consideration the costs of outages."
Not stupid, but MBAs in my experience never actually dig into the spreadsheets and figure out the meaning behind the number. They just see what the vendors promise them over multiple free lunches, golf trips, etc. It doesn't help that most CIOs aren't really technology people, or are so divorced from the day to day operations that they don't know what impact a decision like tha
Re: (Score:3)
The other problem is that most of the decision makers will just bail when the first failure happens, after having collected the bonus for getting rid of the IT team.
And the smart ones bail right after collecting their bonus, before the first failure happens...
darn! (Score:1)
Juuuuust kiiiiiddding. We don't use cloud services.
Office 360 (Score:1)
Simple, just rename it to Office 360.
Problem solved.
Does MS offer any guarantees? (Score:2, Interesting)
I mean, what happens now? If I use Azure in my business, and because this outage I have lost x dollars in business transactions that i could not carry out, is MS going to compensate me in any way? Or is Azure one those services that comes without any guarantees?
Re: (Score:2)
What am I not getting? (Score:3)
I just can't get my head around the idea that somebody would take information vital to their needs and put it beyond reach, under the control of other people whose priorities probably don't match theirs.
What advantages are so overwhelming that they make this a sensible thing to do?
Re: (Score:3)
Cheaper? /annum. We do that on r720 that cost $10k and a couple of hours a month maintainance.
10 x 2core vms is $20k
Uhmmm....What? (Score:2)
Isn't a cloud supposed to be, you know, *distributed*? So that this sort of thing doesn't ever happen, barring a catastrophe of no less than nation-wide proportions, where people are more liable to be more worried about other things than availability of said service anyways.
I would think this is more an illustration of a failure on Microsoft's part to properly implement their cloud services than it is indicative of a failure of such services in general.
Re: (Score:2)
One characteristic of redundant computer systems is that they can all go haywire for the same reason at the same time. Very few other large systems work that way.
Why should "Azure" be any different? (Score:2)
Nothing abnormal here. Sounds like regular Microsoft availability to me.
All Microsoft servers require regular (at least monthly) patching to keep them secure.
All Microsoft products require regular restarting to keep them available and performing correctly when you want them.
Why should "Azure" be any different?
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
You'd think something as big as Microsoft would roll out stuff in phases
You would think Microsoft would do a lot of things the way it "should" be done... but they don't.
Re: (Score:1)
There's a reason Microsoft's address is One Microsoft Way (1 Microsoft Way Redmond, WA 98052).