Microsoft waves goodbye to Small Business Server Small Business Server from Microsoft is "going off the air." With Dell's announcement of unavailability access to the last remaining copies will prove ever more scarce. If you are a managed service provider that specializes in small and medium-sized enterprises (SMEs) I recommend you shake down your suppliers and build up …
If for example you run Wimbledon's website, which has extremely high traffic during the two weeks of the Championships, moderately high traffic in the couple of weeks before it when people are planning their trips, very high traffic when tickets are released for sale, and practically no traffic the rest of the time, then having a cloud service to cope with the demand peaks makes sense rather than buying in the capacity. IBM can then rent out that capacity to Rolland Garros, US Open and so on during the rest of the year.
Hmm - I thought we were talking about managed IT services for SME's. You know - printing, file sharing, firewalls, networking, email provision - all that boring stuff that people *still* need to run a business. Why do some people think that webhosting and internal IT systems are somehow the same?
Cloud assumes a decent Internet connection. A normal 10/10 connection is fine for our 75 staff 400 students. Make that a cloud server, email, storage, productivity and backup solution and we would be ROYALLY screwed. Suddenly not only do we effectively have to pay subscription but also swap our 3k net connection for a 15k one just to replace functionality.
Cloud does not suit everyone but it certainly suits the companies touting it.
Blanket statements are usually wrong.
Cloudy services are probably not a good solution form most companies. But there are some for which it is the ONLY thing that makes sense. For several years I helped a company that had a permanent staff of about 50 people who worked from home, a busy month in which temp staff ballooned to about 300, a customer base of 25,000 to 30,000 people and annual that grew from $1 million to $2 million dollars. Their most important business function took place in the cloud, except we didn't call it the cloud then just an outsourced IT function. The company had an official mail drop for legal correspondence and a different one for business correspondence. Once every other month they'd rent meeting rooms in a hotel to conduct business that required face to face meetings, but otherwise it was conducted by email, phone, and bulletin boards.
Yes, it was a volunteer non-profit, but it was still a business and most certainly an "anyone."
"I wouldn't recommend cloudy service to *anyone*." The use of the word "anyone" in this context is an absolute which might lead a reader to believe your opinion is that nobody should use cloud services.
"Blanket statements are usually wrong." The use of the word "usually" in this context allows for an alternative outcome which might lead a reader to believe that blanket statements are sometimes right - not ironic at all.
Also, opinions can be wrong. Someone may hold the opinion that I am typing this whilst drowning in jam - they would be wrong. You seem to hold the opinion that the phrase "Blanket statements are usually wrong." is ironic - it isn't.
"The use of the word "anyone" in this context is an absolute which might lead a reader to believe your opinion is that nobody should use cloud services."
*Might* lead a reader.... doesn't sound like an absolute to me. BTW, only I know if I'd recommend a cloud or not, so only I can say whether my statement was true or not.
""Blanket statements are usually wrong."
IMHO it is blanket statement because it covers an infinite range of statements. It is also a blanket statement because although responding to *my* statement, It says nothing of whether *my* statement was wrong or not. It's a bit like "weasel-words" in Wikipedia parlance; it tries to devalue without saying anthing very useful for itself. In fact, it is the exception that proves the rule....
""Blanket statements are usually wrong."
give.....
"Blanket statements are sometimes right"
Lol, yeah I read that only a couple of weeks ago.It's close enough for me. Stating that something usually happens, also proves that sometimes the opposite *must* happen. It might not be an exact match, but it is along the same lines. I certainly haven't fallen into the usual mistake, if that is what you are thinking.
Fair enough.
Although it could be better to drop that phrase altogether. It smells funny by now. Edge cases and exceptions are a good way to test validity of the rule, but most of the people just do not bother, it is much easier to declare their assertions correct, and any objections as "exceptions that prove the rule".
> "I wouldn't recommend cloudy service to *anyone*." The use of the word "anyone" in this context is an absolute which might lead a reader to believe your opinion is that nobody should use cloud services.
His action was 'recommending'. 'Not recommending' is not the same as 'recommending not'.
This is a bit of a mixed thing really. On one hand, few things cause more problems than the "special" configurations that were foisted by default on users of Small Business Servers however the cost saving of the Small Business Server bundle compared to the alternatives made them a good solution.
No surprise that MS want to force everything possible onto Sharepoint (they've been beating this drum for years) and their cloud or, more accurately, their subscription offerings.
Agreed. From the bean counter perspective it makes sense. From the customer standpoint it doesn't and for the same reason the bean counters want to change the model.
The thing is, somewhere in the company there's supposed to be someone with the sense to say "We can make X dollars at this lower price or between zero and .5X with a higher priced offering, because people won't buy the higher priced product."
Faced with the current MS roadmap, if I were actively supporting SMEs, I wouldn't be stocking up on disappearing copies of a product MS can kill the support on, I'd be beavering away on alternative solutions.
"Sounds like an opportunity for some Linux disty to put together an 'easy to install/configure' equivilent package."
Unfortunately while unix/linux always had made a far superior back end server to any of the supposed "server" operating systems (I need to reboot a 24/7 server just because I've installed a new driver - seriously??) , they still haven't got to the point where they can be managed by the clueless click monkeys that firms hire as Windows "admins". Unix & linux still need people who have a vague clue about what they're doing and a lot of firms arn't willing to pay for that.
Yep. The components are there: Samba 4 for Active Directory/file sharing, Sogo/OpenChange as an Exchange drop-in replacement, Apache/whatever web server etc.
There are some small business-oriented "all-in-one" distributions like SME Server (former E-Smith), ClearOS (former ClarkConnect) and Zentyal. Haven't looked at them lately - hope they incorporate AD and Exchange functionality now...
We have been using ClearOS for three and a half years now. I don't worry about updates. They just happen. I only reboot it when I have to turn the power off for some reason. I just looked - last reboot was last October. No-one misses Exchange. Some of us use Thunderbird, others Outlook. There's only 8 of us, but for a small business it just works.
To a certain extent, I'd agree, but it really depends on what you want from a server. Many of the distros out there have their benefits and the addition of tools such as Webmin and so forth can really help out the less clued up user. It also depends on the sort of "click-monkey" you refer to as not all Windows "admins" are afraid of getting their hands dirty, though I agree that some would cack themselves if they had to so much as look at a Linux box!
Linux is already easier to install / configure than Windows Server. Problem is that the alternatives to Exchange Server tend to be either more expensive or not that good, and a lot of desktop software requires SQL Server, and while there is no particular reason why it couldn't work on MySQL or something else, it doesn't.
No single reason, but any familiarity with SQL Server and MySQL tells you there are many, many incompatibilities as soon as you move away from anything but a toy database. Transact-SQL has an awful lot of quirks. It is sometimes relatively easy to migrate from MySQL to MSSQL, but not the other way round.
Transact-SQL has an awful lot of quirks.
Some of them are incredibly useful quirks, such as the Hierarchical column data type. Man, that thing makes recursive queries fly by!
Regardless, I don't think SQL Server is by default a part of SBS 2011 (except under the hood) so this isn't all that relevant.
You could replace the non-visible bits with PostgreSQL which does most thing that SQL Server can, albeit in a far less friendly way.
See Trevor Pott's response above.
Yes, the Transact-SQL quirks are very useful, especially the new ones in 2012. SQL Server and Exchange are the reason that people can't migrate to something that ends in x. But I suggest that a hefty dose of realism would persuade anyone that the way the world is going, running them on SBS is not a long term solution regardless of what Microsoft does with the product.
As a user, we moved to Zentyal last year for just such reasons - server over 6 years old.
Really small businesses certainly don't need exchange, but a server and a proper email system is good to have.
It might not be "equivilent " but it is good enough. MS seem to be moving a long way up market.
Thanks for interesting article Trevor.
I used to just be a Windows sysadmin, now I do a lot of Linux work too, and I tell you, Windows is way easier to set up and use, unless you want tot do something particularly out of the ordinary.
On Linux it's almost as easy to set up a really non-standard system, as it is a simple one, or rather, it's just as difficult...
They both have their place.
... Such as SME server. For us, it's a no-brainer. Small company wants a server for mail and file. Windows server means you need to buy MS Office/Outlook and CALs and spend £3-4K on the whole thing. SME server means you probably need to spend £1K. And you can use open source email and office apps. Yes they might not be as refined looking as the Microsoft equivalents but when it costs £180 a pop for MS Office Home and Business 2013, I think most small businesses would put up with the differences.
MS want you to spend money on annual SaaS plans don't they? For most small businesses, MS Word is a tool. They tool works as well the day it's bought as a year later. So why would you want it updated? It's only a word processor for christs sake!!! Companies don't mind paying for antivirus regularly, or backup, because those are things that provide a SERVICE. An ever-changing word processor is NOT a service.
Meanwhile, large American universities which are very influential are abandoning Microsoft products for Google. I suspect that Sharepoint is already over the hill and entering a long phase of slow decline. Azure is a pretty good product but it is one of several in the market, and the notoriously conservative accounting profession is moving at a growing rate to cloud-based back end systems. Why?
The short answer, I suspect, is that people are getting very comfortable with Internet-based systems. All the fears about losing data or losing connection are proving not to be as big a deal as was thought. For most SMEs the on-site infrastructure is a higher risk point of failure than stuff in a secure bunker somewhere that is being monitored constantly by the sort of staff that an SME can only imagine being able to afford, and in the usual critical mass scenario, the more businesses move data and processing offsite, the more cash there is to fund better cloud solutions.
SBS is only good if you have expertise in managing it, and that expertise is going to get more scarce. Which graduate wants a career in something as boring as that, especially as it's perceived it will only have a short shelf life. To use the usual car analogy, the situation now is like the road system in the 1920s. There is still a lot of horse infrastructure, a lot of horse experts, and people are still unsure of how reliable cars will be, and the risks of handing over responsibility to a new breed of mechanics. But now is not a good time to be building a new cart factory, when all the best people want a job with Henry Ford.
"SBS is only good if you have expertise in managing it, and that expertise is going to get more scarce. Which graduate wants a career in something as boring as that,"
I don't think many graduates go into sys admin anyway. Its more the BTEC crowd who fill those jobs. Graduates become coders/designers.
That's basically my point. Do you want your critical infrastructure looked after by someone with a BTEC rather than a graduate? (I mean, yes there are very good BTEC people out there, and apprenticeship schemes which are better than the majority of degrees, but are they going to be administering Exchange and SQL Server 2005 on an SBS box?)
" Do you want your critical infrastructure looked after by someone with a BTEC rather than a graduate?"
I'd like my critical infrastructure looked after by somebody with a clue, not just a piece of paper.
Someone who knows that the way to fix the unknown passwords when the sole sysadmin leaves might well be something like a rescue CD rather than a reinstall from scratch.
Someone who knows that a reinstall *might* be appropriate if the sysadmin's trustworthiness had been known for months to be questionable.
Someone who knows that having a sole sysadmin whose for those servers (in a 2000+ employee outfit) was a pretty damn stupid idea in the first place.
Not a theoretical example.
@ribosome great post, maybe because I am in close agreement. From my view of business cloud take up is growing and in a year will be a flood aand in a few years ubiquitious. Azure will be leading as its underlying infrastructure beats all other Rackspaces or even Google level providers.
Also this move means the role of the system admin is disappearing, just as the small garage owner has gone in your motoring anlogy.
The small branch system admin is most often the most difficult person to get to do good work when rolling out a big system to many locations. Put the system in the cloud and spread it to different centres aound the world and you have a system that can be controlled by specialists in different geographic locations.
All the fears about losing data or losing connection are proving not to be as big a deal as was thought.
For now. Trust me, it's only a matter of time. Wasn't Nintendo just hacked this week? How about the Sony debacle? (3 times in one year?) And didn't web based Outlook go down today? (to be fair, gmail has gone down plenty)
BTEC vs uni grads? I've met plenty of both who were scary, scary incompetent running billion dollar systems and keeping their jobs by successfully blaming everyone else because their boss was an MBA type, who expertise was spreadsheets and therefore they were IT experts, but in reality IT was some kind of black magic voodoo to them.
Back on topic, MS is failing where it really counts: TCO
Why? You're not wanted, support will be knocked on the head, and you'll have to change sooner or later. MS are pushing ahead with the cloud which many people aren't comfortable with. If MS's cloudy offerings turn out to be a commercial disaster then they'll need at least two-three years from now to U-turn.
If they don't want your custom, switch to another solution. There's Novell, RedHat, Oracle, Ubuntu, or Mandriva to name a few paid-for solutions. There's even Apple's point-and-click server. If a free Linux/BSD distribution is acceptable for the company then any non-bleeding edge one which you can support yourself will do.
So why try and stick with MS?
"If they don't want your custom, switch to another solution. "
You clearly have a lot of experience in the field. You clearly have gone through many changes from one system to another, from one vendor's products to another - and we are talking about massive products here - such as an Active Directory/email/sharepoint server - which affect almost every aspect of your client's systems. You have clearly held client's hands (for an entire site) through nightmare upgrades and made sure everybody is happy *and* been able to justify such a torturous path from the point of view of staff training costs, hardware costs, software costs, data conversion costs, business disruption costs etc. to the client.
Yes, I am being sarcastic - but if you really care about your clients, about their business efficiency and about helping them make the most of their IT systems - you would know what I mean. Then again - some people only care about the sales and the commission they get - the rest is not their problem.
I am not a fan of MS products and use them as little as possible - but even I wouldn't advise a wholesale change from an MS set of backends to something else of my preference unless there simply is no other choice.
We're not talking about wholesale change, we're talking about buying new copies of SBS for new clients or to be used to roll out additional servers for existing clients. If the client is new then it doesn't have to be MS. If the client is existing then one non-MS server alongside the rest won't bring the house down. Those clients being small businesses where any change is less disruptive... which is why MS thinks they can bully them all onto the cloud in the first place.
If you really care about your clients you'll sit down with them and explain the current situation. Yes the switch to a non-MS product will probably be difficult for them and maybe you. But if MS wants to treat them through you like red-headed step children it is probably better for both of you to deal with that issue up front. Because support for you and your client isn't going to get any better if they opt to follow the MS cloudy route.
And you'll be the one the customer berates because MS isn't supporting you the vendor in the manner required.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
