Microsoft Announces SQL Server On Linux

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
Microsoft announced today (via this blog post) that it is bringing a preview of SQL Server to Linux. The company says it has plans to make the product generally available by mid-2017.

Today I’m excited to announce our plans to bring SQL Server to Linux as well. This will enable SQL Server to deliver a consistent data platform across Windows Server and Linux, as well as on-premises and cloud. We are bringing the core relational database capabilities to preview today, and are targeting availability in mid-2017. SQL Server on Linux will provide customers with even more flexibility in their data solution. One with mission-critical performance, industry-leading TCO, best-in-class security, and hybrid cloud innovations – like Stretch Database which lets customers access their data on-premises and in the cloud whenever they want at low cost – all built in.
 
Nothing wrong with this as long as it works correctly. However, it is on a server base and probably would be installed on an LTS Linux install so it makes sense that it would be stable.
 
"I give up" - Microsoft 3/7/2016
Nonsense, they are a software company. This release should have happened as soon as a large enough customer base materialized to be profitable for them.
 
They have had Linux support for Years for other applications/products, this should be more like about time. I like windows, but if you cannot look at ms or open source like tools, and use the right tool when necessary, then you'll fail.
 
Nonsense, they are a software company. This release should have happened as soon as a large enough customer base materialized to be profitable for them.

The question is... Who is the market for this ?

The answer is companies that have been getting rid of their MS stuff... or newer startups that never had any to begin with. MS has been loosing mind share in that market for so long that not many companies see the value of trying to mess around with windows based SQL and even though they started selling linux versions of their cloud stuff a year ago, the adaptation isn't likely where MS wants to see Azure. This may help them get into a few larger data centers and make their cloud options more attractive as they may be easier to work with if its possible to run SQL on their existing linux powered setups. I would guess this isn't going to go as well as MS hopes, they will need a few big wins to help push mind share. I would expect MS will go as far as GIVING SQL to some big partners and Giving them a lot of free support to make it happen. To make real inroads in that sector they need some big mind share earning wins. I think its still going to be a real uphill battle to win any larger customers for them even if they are literally paying someone to implement it... unless MS is willing to open source it. Right now one big advantage of the open options is being able to slice and dice and make it do things you need like run on OSx in offline mode for devs ect... not sure how open the biggest users are going to be to something completely closed off, and I'm not sure MS is really ready to opensource a profitable software.
 
Last edited:
The question is... Who is the market for this ?

The answer is companies that have been getting rid of their MS stuff... or newer startups that never had any to begin with. MS has been loosing mind share in that market for so long that not many companies see the value of trying to mess around with windows based SQL and even though they started selling linux versions of their cloud stuff a year ago, the adaptation isn't likely where MS wants to see Azure. This may help them get into a few larger data centers and make their cloud options more attractive as they may be easier to work with if its possible to run SQL on their existing linux powered setups. I would guess this isn't going to go as well as MS hopes, they will need a few big wins to help push mind share. I would expect MS will go as far as GIVING SQL to some big partners and Giving them a lot of free support to make it happen. To make real inroads in that sector they need some big mind share earning wins. I think its still going to be a real uphill battle to win any larger customers for them even if they are literally paying someone to implement it... unless MS is willing to open source it. Right now one big advantage of the open options is being able to slice and dice and make it do things you need like run on OSx in offline mode for devs ect... not sure how open the biggest users are going to be to something completely closed off, and I'm not sure MS is really ready to opensource a profitable software.

Um, no. More like a cost effective replacement for Oracle. MS is going to pull some rug out from under Oracle.

Besides, like others have stated, they are a software company. There is money to be made in Linux.
 
Um, no. More like a cost effective replacement for Oracle. MS is going to pull some rug out from under Oracle.

Besides, like others have stated, they are a software company. There is money to be made in Linux.

Bingo. This is meant to compete with a other commercial RDBMS in the enterprise, Oracle being a big one along with IBM's Neteeza. SQL Server is very popular in the enterprise and this will only help as Linux servers are big in the enterprise as well. And yes, Microsoft is a software company. It's become pretty open to developing on platforms where it makes sense and this makes perfect sense.
 
I wonder how much this would effect the market for MySQL. The idea of being able to run MS SQL on a more reliable OS than Windows Server is interesting, assuming they make all the functions and syntax that work on the MS Server version also work the same way on the Linux version.
 
I wonder how much this would effect the market for MySQL. The idea of being able to run MS SQL on a more reliable OS than Windows Server is interesting, assuming they make all the functions and syntax that work on the MS Server version also work the same way on the Linux version.

It's a given that the Windows and Linux versions are going to be identical at the application layer, that's always been the history of cross platform RDBMS, you don't care much about the underlying platform. Not sure why some people seem to be freaking out over this. This is really good thing overall, especially if you're a long time SQL Server developer like me. It's going to bump a few Oracle and Neteeza enterprise installs out there I think without question. Maybe not so great from a Windows Server perspective but there's tons on Linux servers in the enterprise. That's not going away and neither is Windows. There's advantages and disadvantages to both. In my organization, there's a lot of love for SQL Server and there's going to be people there very happy with this.
 
It's a given that the Windows and Linux versions are going to be identical at the application layer, that's always been the history of cross platform RDBMS, you don't care much about the underlying platform. Not sure why some people seem to be freaking out over this.


When it comes to Microsoft developing anything for another platform, even if it's a different version of their own OS's, nothing can be taken for granted. If you've ever tried something even so simple as importing an HTML signature in Outlook from a Windows machine to Outlook on a Mac, you know exactly what I'm talking about. When it comes to developing software, I've always had the impression MS has way too many cooks in the kitchen, and none of them communicate properly. I think I recall a story quite a while back that they have a bunch of teams and they don't work with each other as well as they should or could. While it makes sense that all should be as expected, I never underestimate any big company's ability to fuck shit up.

On a side note, I wonder if this means that SQL Express would be Linux-compatible some day. That would be really interesting.
 
When it comes to Microsoft developing anything for another platform, even if it's a different version of their own OS's, nothing can be taken for granted. If you've ever tried something even so simple as importing an HTML signature in Outlook from a Windows machine to Outlook on a Mac, you know exactly what I'm talking about. When it comes to developing software, I've always had the impression MS has way too many cooks in the kitchen, and none of them communicate properly. I think I recall a story quite a while back that they have a bunch of teams and they don't work with each other as well as they should or could. While it makes sense that all should be as expected, I never underestimate any big company's ability to fuck shit up.

On a side note, I wonder if this means that SQL Express would be Linux-compatible some day. That would be really interesting.


I'm hoping they make express available for Linux as well. I like postgresql but Microsoft has better management tools. Most of my use cases wouldn't hit the express limits.
 
When it comes to Microsoft developing anything for another platform, even if it's a different version of their own OS's, nothing can be taken for granted. If you've ever tried something even so simple as importing an HTML signature in Outlook from a Windows machine to Outlook on a Mac, you know exactly what I'm talking about. When it comes to developing software, I've always had the impression MS has way too many cooks in the kitchen, and none of them communicate properly. I think I recall a story quite a while back that they have a bunch of teams and they don't work with each other as well as they should or could. While it makes sense that all should be as expected, I never underestimate any big company's ability to fuck shit up.

On a side note, I wonder if this means that SQL Express would be Linux-compatible some day. That would be really interesting.

There are 5 different native platforms that parts of the Office suite support, Win32, Universal Windows, OS X, Android and iOS. That kind of requires a lot of cooks and there's always going to be at least minor variances in functionality. But overall Office does a pretty good job of being cross platform. It will be imperative for SQL Server to function nearly identically, there may be some more esoteric features and capabilities that aren't, but I'd expect close to 100% interoperability at the application layer.

An express version for Linux would also be great but this is really for competing against commercial software, there's not the incentive to support the express version though I'm guessing that'll happen sooner or latter. I could see a lot of folks using FOSS development tools pick up SQL Server Express because it is very good, at least on Windows, for what it is.
 
The Tide is turning and this is just a incoming wave.

As mentioned, a lot of new companies are not going with MS products since open source software has proven itself especially for servers.
I have to give props to Microsoft for the all encompassing monopoly they had for 25 years. Very shrewd business (if not somewhat crooked) This ranged from strong arming PC manufacturers and VARS to only allow MS windows to be preloaded (No choice for no OS even). To the IT training and the huge PUSH 15 years ago for every IT and computer person to wave a MCSE certificate around thus insuring that according to them (run of the mill IT people) there is no alternative to Microsoft products. All this is slowly coming to an end.
 
The Tide is turning and this is just a incoming wave.

As mentioned, a lot of new companies are not going with MS products since open source software has proven itself especially for servers.
I have to give props to Microsoft for the all encompassing monopoly they had for 25 years. Very shrewd business (if not somewhat crooked) This ranged from strong arming PC manufacturers and VARS to only allow MS windows to be preloaded (No choice for no OS even). To the IT training and the huge PUSH 15 years ago for every IT and computer person to wave a MCSE certificate around thus insuring that according to them (run of the mill IT people) there is no alternative to Microsoft products. All this is slowly coming to an end.

This has little to do with any Tide turning, unless you are referring to IaaS taking over the datacenter and the OS being abstracted completely in the next 5 years? This is more about MS positioning themselves w/Azure and AWS. They do won't need Windows Server as an OS in the next 10 years. This is no longer a Windows vs Linux debate, but making sure their back-end software as a whole will run on whatever IaaS/PaaS/SaaS that is out there. Because, THIS is where the industry is going.
 
SQL Server is bonkers expensive, not sure who would pay for it, and be a Linux OS user.
 
Not compared to Oracle and Neteeza.

Cost wise they are pretty much on par up front for sure. Oracle might even be a bit more expensive. For the largest players though being able to work directly with Oracle and tweak stuff is still likely going to be a big plus for them over MS. I still don't see MS being open enough to please some of the larger data centers that have been used to being able to see touch and modify pretty much anything as needed. Will be interesting to see the numbers by the end of 2017 (assuming they don't launch it in Nov or something). The market is pretty large, I have no doubt MS will be able to make a bit of headway, I am just curious if they will be happy picking up the medium size end of the business.
 
Cost wise they are pretty much on par up front for sure. Oracle might even be a bit more expensive. For the largest players though being able to work directly with Oracle and tweak stuff is still likely going to be a big plus for them over MS.

I work in the type of environment that SQL Server Linux will appeal to. Microsoft works with us on issues and there are places I know at the bank who will like this option, assuming the quality is a good or better as SQL Server on Windows.
 
Cost wise they are pretty much on par up front for sure. Oracle might even be a bit more expensive..

Have... have you ever priced it out? Oracle is sometimes an order of magnitude more expensive.

(Assuming Xeon E5-26xx)
Per-core pricing on SQL 2014 Ent is around 7500 list, for DB EE it's $23750. Oh, don't forget the Oracle tuning ($2500/c) and diagnostics packs ($3750/c) , partitioning ($5750/c), RAC($11500/c), data compression($5750/c) , OLAP($11500/c), etc etc.
 
I work in the type of environment that SQL Server Linux will appeal to. Microsoft works with us on issues and there are places I know at the bank who will like this option, assuming the quality is a good or better as SQL Server on Windows.

Have... have you ever priced it out? Oracle is sometimes an order of magnitude more expensive.

(Assuming Xeon E5-26xx)
Per-core pricing on SQL 2014 Ent is around 7500 list, for DB EE it's $23750. Oh, don't forget the Oracle tuning ($2500/c) and diagnostics packs ($3750/c) , partitioning ($5750/c), RAC($11500/c), data compression($5750/c) , OLAP($11500/c), etc etc.

I agree there is a market for MS to go after. I honestly give them a nod for having the guts to chase it down and go after it, considering the possible optics of supporting the competitions (even if its opensourced) OS. The market is in those medium size outfits no doubt... I doubt MS gets any traction with the largest database users... the market for all those medium size operations though is of course very large. Like I have said earlier it will be interesting to see how this shakes out for MS... are they going to go after a few of the big guys and offer to almost pay them for the PR... or do they quietly go about taking as much of the legit market, which yes I believe as well is the middle size Large enough but not massive size operations like most of the smaller banks in the US ect.

I don't mean to keep sounding so negative on MS... but imo if they go after a big datacenter or something in a way where you can tell they are obviously throwing out tons of free support and perhaps even kicking back to get it done, then I say MS knows they have lost the market and are just putting on a show for investors more then seriously chasing the market. Should be good entertainment anyway.
 
Back
Top