Will Oracle kill MySQL? Who cares?

When Oracle bought Sun last week, the MySQL community collectively curled its lip into a worried sneer. Is Oracle going to kill MySQL? We'll have to wait and see. Oracle's acquisition of MySQL isn't like a lottery ticket. There is no chance that MySQL will get significantly better by leveraging some of Oracle's technology. …

[amanfromMars]

One Small Step for a Man .....

"Larry Ellison understands that a small, vocal minority thinks they control the future of databases, but real-world business pragmatism is slow to change paths"

Ted,

It would be interesting/enlightening to know if Larry Ellison thinks that thought to be true or false/perfectly reasonable or hopelessly delusional........ although I would imagine them to a QuITE quiet thoughtful bunch rather than a small, vocal minority, although they would be able to be both at Will.

Fortunately, it is not slow real-world business pragmatism which changes paths, but virtually real entrepreneurs into changing paths..... Shining PathFinders.

And as for databases, the One which Contains and can Present Tomorrows' Information Today are the Really Powerful Ones ..... and they would be Virtual MetaDataBases. Does Larry have any of those Churning Bits and Bytes in Butter for the Oracle Operating System?

[Chris C]

Yes and no

"What software managers recognize - and often need to beat into their engineers - is that it is far easier and cheaper to throw money at a scalability problem than it is to throw engineering resources at it. In the case of web service providers using MySQL, when they hit a scalability wall, Oracle can now step in and sell them a database and Sun hardware that will fix the problem, no re-architecting required."

Part of that is correct -- sometimes it is easier and cheaper to throw money at a scalability problem. However, like any problem, proper engineering and optimization may solve the problem and save money long-term. That's where those managers fail -- recognizing and acknowledging long-term effects. Managers, CEOs, etc only seem to care about the short-term, right now, and don't even consider the long-term effects (whether it's software, hardware, or any other industry).

However, I wouldn't be so quick to say that Oracle can sell a database and Sun hardware to a MySQL user that will fix the problem with no re-architecting required. After all, MySQL isn't exactly standardized; it's like Internet Explorer. They took a standard and extended it how they saw fit. It includes a number of MySQL-only functions which are not part of any SQL standard. If any of those proprietary functions are used, re-architecting WILL be required. That isn't saying that it would or wouldn't be cheaper than re-engineering the web app, but it is a reality that needs to be considered. Far too often, salespeople say one thing (such as "you can just replace your DB with ours, no changes needed"), and the engineers are left to get it working when it turns out that it was misrepresented.

[Gerhard Mack]

MYSQL is already being replaced

SQlite is much better for web serving than MySQL is and PostgreSQL is better on the high end. Unless you need replication your better off with one of the other two.

[Ian Bradshaw]

XE ...

Why not just use XE? MySQL is pitty much legacy for existing sites ever since Oracle gave us XE for free.

All Oracle need to do is raise the XE limit from 2GB storage and job done. It's Oracle mini, which is all MySQL is. You can't store massive amounts of data in MySQL, and with decent table spec (such as using the right data types and lengths for example), its amazing how much you can actually cram into 2GB. Short of space? Just create another copy of XE in VMWare shell and get yourself some clustering experiance (the hard way as it's the only Oracle feature not in XE that I've come accross, so you need to think a bit more - no bad thing).

Give it a bit more storage in 11 XE and then we can all use that and truely kill MySQL, learn some Oracle skillz and migrate any apps we do to full scale corporate at some point with no worries about chaning quieries.

XE is the Oracle version of a community server (i.e. free for whatever), and has a lot more features.

Just because it's not open source, doesn't mean it's not better and provides an easier expansion to corporate land as well as giving the new developers some good Oracle experiance.

Oracle killed MySQL long before they bought it, they killed it as soon as they released XE.

[Anonymous Coward]

Best bit

The article was not encouraging reading for those who want MySQL to thrive. However, the best sentence was:

"Oracle's open source strategy resembles a sixteen year old boy putting his complete mental capacity to work trying to figure out how to get high off of household chemicals".

That made my Monday morning a bit brighter.

[Anonymous Coward]

Woot: You don't say...

"From what I can gather, Unbreakable is nothing more than a marketing campaign around a Linux support program, the main selling points of which are:"

Hallelujah! That's all *every* linux distribution is; some happen to be presented better than others

As for MySqL, what's the problem? It's source code is in the open domain and it will continue to be so. Or are you implying that the various open source licenses can be revoked? And exactly how would that work? How would you retrieve *all* the copies that are out there sitting on tape backups?

Besides, I've always preferred Postgres

.

[jake]

Larry!san can't teach much of anything.

"Now, Ellison is teaching it fear."

I'm shaking in my boots. Not.

[Anonymous Coward]

A herd of elephants

> PostgreSQL is better on the high end. Unless you need replication your better off with one of the other two.

Someone clearly forgot to tell Skype that Postgres doesnt replicate or scale...

(https://developer.skype.com/SkypeGarage/DbProjects/SkyTools)

They seem to be happily running their business using it...

[Anonymous Coward]

@Gerhard

lol ... your a joke. Not that I dislike PostgreSQL, but it had its chance, now it stuck in the gutter with the select few hardcore fan boys.

Yes, they may kill the current the MySQL "Enterprise", but its to late to destroy the community edition.

[Dave Gomm]

They don't actually have to do anything

I might be well off the track here but perhaps the most interesting strategy for Oracle would be to actually do nothing, instead they could just make lots of noise about how much they intend to do to support and extend MY SQL (and promote FUD via unofficial channels) - what better way to confuse and distract the open source community and their business clients, keeping it from being credible in the enterprise but at the same time using it to stifle any real open source opposition.

[John Latham]

It depends

As usual, there is no one answer. If there was, everyone would use the same product.

MySQL is fine for many read-heavy web workloads, where application-level caching can do a lot of good. What proportion of websites will ever get more than a few hundred page impressions a second?

What annoys most engineers is being told that a product which works perfectly fine for them is a heap of shit, and that they should instead use some newer/shinier/more expensive piece of shit instead.

Personally, if I was architecting a new website for massive loads I'd seriously consider something like Google App Engine (or at least something that was easily portable to it). No doubt there are lots of people who would tell me this is a stupid idea.

YMMV, whatever works for you.

[Anonymous Coward]

@J Latham

Childish moment, but please allow me to take the following line out of context:

"architecting a new website for massive loads"

Ooooo, matron!

[Tony Hoyle]

MySQL killed themselves

By GPLing their client library they made it impossible for anyone to use it unless they were producing GPL software, or they paid for a license for every client at £200 a copy. No other database imposes such limititions merely on the client libraries.. they license the number of users to the server.

We tried to negotiate with MySQL - pointing out that was four times the retail of the entire product and if you included the free editions (which they insisted that we do) that would come to a payment something like the national debt of Brazil. They wouldn't budge.. that's their price and they're sticking to it... so we dropped support for them. We're not alone in being caught by this.

No such problems with Oracle, MSSQL, Postgres, etc. They realize that trying to stop developers supporting your product is counterproductive.

[Mark Rendle]

Know what works really well for web serving?

SQL Server 2008.

[Anonymous Coward]

I don't like Mondays

But I'm glad Ted is back. I was beginning to wonder if he had run out of spleen to vent.

[adrian lynch]

@Evil Graham - Welcome back Ted

Apparently he was having a baby. Can I be the second to say what a pleasure it is to have Ted back on the Reg.

[Eponymous Cowherd]

Nice to see that.....

El Reg's pet troll is still alive and well.

[Anonymous Coward]

Ted!

"Good to see you!

I still don't understand a word of what you say, but it sounds very good, and very funny.

Best line this time? The car keys/pants one. Hands down (fnar) winner.

Congrats on the little Dzuiba. Although, if he's anything like his dad, I doubt he'll sleep through the night. Much wailing and screaming to come, I think."

[Dave]

@Tony Hoyle

What you say stands rather contrary to my understanding of the GPL - if you want to include a library or other complete component in you project then you are free (in both senses) to do so.

It is only when you start modifying those components, or using small pieces of code that you have any real problem.

[Anonymous Coward]

Hmmmm

------------------

With the attention span of a fruit fly, your average web programmer is quick to latch on to new technologies. After all, as a web developer, you can't appear publicly to be older than thirty.

------------------

I'm 32 - I can has pension? kthnxbye

Still - with Oracle owning MySQL now - I'm glad we've got a nice stable database factory wrapper underneath our site. A db server switch is a relatively painless operation these days.

However, I'll not be switching yet - I'm, at worst, ambivalent towards Oracle and considering Sun has lacked direction with, well everything, over the last few years the Oracle acquisition might not be a bad thing. I can foresee enterprise MySQL being killed off but overall the buy might enable Oracle to gain some leverage in the web and SME "theatres of conflict" - and for that they might focus (and maybe improve) MySQL.

However, I smell Oracle MySQL support contracts on the horizon... and they'll not be cheap.

[jon]

swearing

is it really necessary to swear? It turned a mediocre article into a waste-of-my-time-reading-it article.

Also, ditch the gamer speak as it proves your IQ.

kthx bye o/

[TimNevins]

@Ian Bradshaw (XE)

Ian,

Unfortunately you cannot run XE using VMWare Shell as this is prohibited.

Not only that but if you have more than one CPU it will only use just the single CPU.

So it is not as beneficial as it first seems.

http://www.oracle.com/technology/software/htdocs/xe_lic_prod.html

"Any use of the Oracle Database Express Edition is subject to the following limitations;

1. Express Edition is limited to a single instance on any server;

2. Express Edition may be installed on a multiple CPU server, but may only be executed on one processor in any server;

3. Express Edition may only be used to support up to 4GB of user data (not including Express Edition system data);

4. Express Edition may use up to 1 GB RAM of available memory."

[Anonymous Coward]

@Mark Rendle Posted Monday 27th April 2009 09:57 GMT

>Know what works really well for web serving?

>SQL Server 2008.

Oh, I needed that. A grey dreary Monday morning with little hope of let-up at work, and someone tells me a fantastic joke to cheer me up!

I'll clean the coffee off my screen and go back to migrating last weeks mess from SQL Server 2008 to a proper database server...

Thanks Mark :-)

[Anonymous Coward]

@ jon

You must be new here.

NOOB

[Niall]

Good article

That is all

[Brian Reich]

Great Article!

Its rare that I find a tech article with even an ounce of humor, but this one was both informative and made me chuckle, especially with the 16 year-old gas-huffer metaphor :)

The merger of Oracle and SUN makes me very pleased with myself for A) having adopted PostgreSQL for many of my projects, and B) using database abstraction where I can. If MySQL does fall by the wayside I'll be just fine. I just worry for the thousands of users out there using free software that is 100% dependant on MySQL, such as WordPress.

[Anonymous Coward]

@Dave

Actually, I believe the original posting by Tony Hoyle is correct. LGPL was invented to provide library writers with a mechanism for writing libraries that could be used without having to GPL the code that used it. But LGPL was largely discouraged and most people started issuing libraries that were GPLed. This forces someone using the library to issue under GPL and has stopped me using a whole bunch of good libraries. On a few occasions I've had to basically write a library from scratch because I'm not going to make highly proprietary algorithms open source.

[Anonymous Coward]

@jon

Perhaps El Reg could accomodate your request by introducing a new acronym for such articles. For example, they already use NSFW ("Not Safe For Work") for those itemswhere there are pictures of boobies, or other rude things that might cause embarrassment in an office environment.

Might I suggest LOFS ("Loads of Fucking Swearing") as a suitable disclaimer? Then you can easily avoid them and - as an added bonus - I can read those ones first.

Job done, everyone happy.

[A J Stiles]

@ Tony Hoyle

"By GPLing their client library they made it impossible for anyone to use it unless they were producing GPL software, or they paid for a license for every client at £200 a copy."

What's the problem? You wouldn't have had to pay a single penny for licences, if you just released your software under the GPL.

"We tried to negotiate with MySQL - pointing out that was four times the retail of the entire product ..... They wouldn't budge"

No; because the whole freaking idea of the GPL is to encourage you to release your software under the GPL. It's called reciprocity; if you benefitted from the hard work of others, then others deserve to benefit from your hard work. Evidently, you preferred to spit your dummy out of your pram. And now, people looking specifically for GPL applications will have one fewer choice to consider.

[drag]

GPL is a problem for some. That is what made MySQL profitable, silly

The problem is that they were probably hoping to make money from their software.

Apparently the only way some people can figure out to make money is by using copyright law to restrict supply and then profit from any potential demand. Works for Oracle, Works for Microsoft.

GPL precludes that approach.

But......

What happenned to MySQL is the same thing that happens to any Open Source project that Sun has control over. Sun gets all stuck up, pisses off people that want to work for them, for free, and drives them away from contributing to said project.

Then they go and throw a few 'real' engineers at the problem and then releases some substandard peice of shit. The MySQL 5.1 release immediately put people on edge.... The 5.4 isn't helping anybody.

That is why you have things like Drizzle, MariaDB, and OurDelta. That sort of thing is a response to Sun's governance of MySQL. I mean you didn't have those forks and branches before Sun took over and MySQL was actually profitable and produced decent enough software.

The same thing is happenned with OpenOffice.org. Sun Microsystems refuses to accept fixes, patches, improvements, and other things becuase.. fuck if I know. They just don't work well with others.

That is why you have things like Go-OO.org. Thats generally the actual OpenOffice that Linux distributions ship. With that you have performance improvements, better document compatibility, better VBA script compatibility, and lots of other improvements, bug fixes, and features.

I have a feeling that Oracle may get some very bad Sun flu pretty soon...

[Anonymous Coward]

@Stiles

>>"We tried to negotiate with MySQL - pointing out that was four times the retail of the entire >>product ..... They wouldn't budge"

>No; because the whole freaking idea of the GPL is to encourage you to release your software >under the GPL. It's called reciprocity; if you benefitted from the hard work of others, then others >deserve to benefit from your hard work. Evidently, you preferred to spit your dummy out of your >pram. And now, people looking specifically for GPL applications will have one fewer choice to >consider.

Fuck 'em. They weren't going to get another GPL program anyway - sounds like this was a proprietary project. They tried to bargain, MySQL played harball and lost. If the end user wants free software they'll just have to suck it up and use what's there. But since all OSS stuff is automatically better, that's cool, right? The point you're missing is that the guy was willing to pay, just not that much. No dummy-throwing, just business. Some people write programs to pay their bills, after all. I know gimps like you think you can make a living from support when everyone's on OSS, but it really doesn't strike me as a sound model. Especially when the standard response when someone has a problem with something goes 'RTFM, noob' -> 'Well, it's better than M$' -> 'You can't complain - it didn't cost you anything' -> 'I'd like to see you do better' -> 'If you're so stupid fuck off back to M$ then.' Usually there's also an accusation of piracy if anyone's using proprietary stuff in there too. And you and your ilk accuse other people of childishness? What a maroon ...

[Chris C]

XE?

"Why not just use XE? MySQL is pitty much legacy for existing sites ever since Oracle gave us XE for free... Just because it's not open source, doesn't mean it's not better and provides an easier expansion to corporate land as well as giving the new developers some good Oracle experiance."

You really don't "get" open source, do you? One of its basic tenets is that, having the source source, you always have the ability to support the software yourself (or have someone else support it for you). For example, when the author decides they no longer want to support their software. Is this used often? Probably not. But it certainly is one "selling point" of open source software.

Having said that, I think MySQL killed itself when they switched to an annul commercial pricing plan in which you have a recurring per-server annual fee. Also, reading their MySQL Enterprise page, it's good to see they've joined the ranks of the idiots who still haven't figured out that "unlimited" means "without limit". The features page ("http://www.mysql.com/products/enterprise/features.html") has this wonderful footnote: "1. Consultative Support: Max 8 hours/server/year Enterprise, Max 100 hours/year for Unlimited. " I bet you didn't know that "unlimited" meant "100".

--------------------

As for GPL software, there are many software apps (I would guess it's more like most software apps) don't need support. As such, an author who releases their product under the GPL will, in all likelihood, not receive anything for their product. In other words, they'll never recoup the cost it took to create the software. Effectively, the GPL is only beneficial to those projects which will generate substantial support revenue, and those projects worked on by people in their spare time (and, of course, people who simply don't want to pay for software). Anyone who wants to make a living writing software has almost no chance doing so with the GPL unless they can convince a lot of people to purchase support contracts (at which point they most likely stop being a software writer and start being a help desk operator).

--------------------

@Dave: "...if you want to include a library or other complete component in you project then you are free (in both senses) to do so..."

That is incorrect. From GNU's FAQ ("http://www.gnu.org/licenses/gpl-faq.html#IfLibraryIsGPL"):

"If a library is released under the GPL (not the LGPL), does that mean that any program which uses it has to be under the GPL or a GPL-compatible license?

Yes, because the program as it is actually run includes the library."

[Chris C]

@AC 17:05

"Especially when the standard response when someone has a problem with something goes 'RTFM, noob' -> 'Well, it's better than M$' -> 'You can't complain - it didn't cost you anything' -> 'I'd like to see you do better' -> 'If you're so stupid fuck off back to M$ then.'"

That sounds about right. I particularly like the "you can't complain - it didn't cost you anything" line, as if the price you pay has any bearing on what you can say about it. Fine, but it goes both ways. If zero-cost means you can't criticize a product (or even recommend new features), then it also means you can't praise that product.

Also, there's one response you forgot: "That's already been discussed. Search for it". Of course, you usually find those answers by doing a search, and usually are still finding those answers after hours of searching. Open source would be a lot more successful (and, I'd wager, generate a lot more revenue) if "the community" didn't have such an elitist and arrogant attitude. Yes, *I* can compile and install from source when I see a new software app I want to try out, but my mother sure won't be able to.

[Anonymous Coward]

Yes frankly who cares

MySQL was used because it was offered on ISPs, so if you are MySQL DB thingy, it is advertising the fact you don't really know much about databases.

SQLite and PostgreSQL those are the ones.

[Anonymous Coward]

@Chris C

> "You really don't "get" open source, do you? One of its basic tenets is that, having the

> source source, you always have the ability to support the software yourself (or have

> someone else support it for you)."

And when last did you a source code update to mySQL, Chris? Do you think that John Doe and the remainder of the the vast majority of mySQL users care whether or not they can hack mySQL code?

What you don't "get" is that 99.999% of users and business do not give a rat's ass about the Open Source "philosophy". They want software that is stable, robust, does the job, at a minimal cost. If Open Source fits that, great. If not, no amount of open source religion spouted by fanbois will get them to switch to Open Source.

Oracle XE (eXpress Edition) is just that. Despite its limits of db size and resource utilisation, it just works and is free. And works damn well. And sports a web browser IDE (aka APEX) that can get even John Doe to create a workable web site with 0% knowledge of web architecture, what the deal with stateless web sessions is, how to deal with optimistic locking, how to manage web state, cookies, authentication and the rest.

And it is for this very reason while it took for someone like Shuttleworth to get Linux sold in the minds of end-users as as a viable desktop solution. It is not about the Open Source as a Religion. It is about Open Source as Robust Working Solutions.

Mine is the one with Screw Open Source Zealots in the one pocket and Open Source As Real World Solutions in the other.

[Doug Southworth]

I'm a little dissapointed...

"After all, as a web developer, you can't appear publicly to be older than thirty."

Ted, no Logan's Run reference? Tisk...

[Don Mitchell]

Amdahl's Law

It's tough to get any objective data, since benchmarking is so controversial in the open source world, especially the industry standard TPC-C. They really want you to look at the ISAM engine performance. It's considered FUD to even suggest comparing MySQL to the commercial enterprise databases, but people have done it. Lately, InnoDB runs that benchmark at about 1/3 the speed of Oracle/Sybase/MS SQL Server. That's a big improvement since Dafodil ran the tests several years ago, and InnoDB was 1/50 as fast then!

The real problem is multicore performance, where the commercial systems leave MySQL in their dust. I don't think MySQL can easily jump that hurdle, because it involves deep and sophisticated architectural changes to push against Amdahl's Law. The Linux kernel itself levels off at 4 cores, due to various global resources and spin locks (Windows NT kernel hits the wall at 8 cores or the same reasons). SQL Server as an application runs well on up to 128 cores, thanks to the SequelOS layer -- and in the new Red Dog operating system, that multicore performance has been absorbed into the OS kernel itself.

There are some big challenges ahead for Linux and MySQL if they want to play in that arena, because multicore hardware is only going to get more "multi".

[Francis Fish]

It's NOT plain sailing to move from one to the other

create table x ( x varchar(2) ) ;

insert into x values('X') ;

select * from x where x = 'x' ;

The select will return different things on Oracle and MySQL - mysql ignores case and Oracle doesn't. I've no idea how many million apps this will break. The Oracle solution is technically correct. SqlLite is case sensitive too.

Also, you can store an empty string in MySQL - oracle always stores null.

select * from something where thing = ''

Will never find anything on an Oracle database, but will find stuff on MySQL if empty strings have been stored.

LOTS of stuff will be broken. Tee hee.

[Anonymous Coward]

Hold On a minute...

Forgive me, did I finally understand one of amanfrommars posts?

I kinda could tell he's a smart guy after dissecting the many earlier posts but

missing much, going away feeling I wished I understood what point

he was trying to make.

knowing there was one deep within......

Paris cuz she's a deep thinker and might post a translation of amanfrommars's posts?

[Anonymous Coward]

Oracle mini vs. the GPL

@Ian Bradshaw

<blockquote>It's Oracle mini, which is all MySQL is.</blockquote>

Oh really, it's Free software, is it? Believe it or not, for some of us that is a required feature. (Yes I have a real employer; we use MSSS in our 4000-server production infrastructure, but we also use a lot of MySQL (along with PostgreSQL and some sqlite in places) in middleware, system management, monitoring, end-user UI web apps,.. )

[Anonymous Coward]

MySQL: What is it good for?

I can see how multicore performance may not measure up to the big boys, but...

Who cares? If a product doesn't fit your needs, use a different one, don't waste time slagging it off, especially if you didn't pay anything for it - you weren't cheated out of anything! It isn't a question of "what did you expect for free?" its a question of "one size doesn't fit all."

There are plenty of applications that don't need multicore performance accessing TB of data with a zillion queries per second. Sometimes all you need is something easy, quick and resource-light. My mythtv config still uses mysql and it works fine. There's no need for Oracle for that.

As for architecting (is that even a word?) around the scalability issues there may be significant benefit to sharding with mysql or postgres or whatever on x86, over going with Oracle. It isn't a question of "working around" the problem of scaling, its a question of building scalability into the design. From a networking background, its a bit like either buying a 10gb/s ethernet link or aggregating 10 1gb/s links. Certainly the 10gb/s link would be cool and a neater solution, but its very expensive and what if you need 11gb/s or only 5gb/s? You still need a "horizontal" scaling design. The problem with Oracle is the licenses are expensive so you try not to scale horizontally. Open source software allows you to add more cheap, simple application servers rather than using expensive and complicated software/hardware engineering to solve the problem in a single box.

AC because, well, it appears DB guys are a bit aggressive!

[Anonymous Coward]

MySQL Must Stay

It will stay if only to irritate MicroSoft. Remeber--- MicroSoft has strategically targeted small databases since it started trying to write operating systems. InterOffice Mail proved a more viable vector for them and they have since added small web servers. But small databases are still a high priority for them.

[asdf]

friggin oracle shills

To all the Oracle paid shills pushing Oracle XE know that developers can smell the cheap cologne of hack salesman a mile away. You have some valid points about open source but Oracle's business model is legal extortion. I don't want to use their great "free" product that is meant to pidgeon hole me into spending way too much in BS oracle shit later. I guess you country club Republican salesmen really hate that "commie" open source stuff. Fear is good.