Oracle Lines Up Unbreakable MySQL

munchola writes "MySQL CEO, Marten Mickos, has revealed to CBRonline that Oracle has threatened to provide support for MySQL and is already distributing the open source database. "They have hinted to us that they will," said Mickos, indicating that the database giant is planning to repeat its October 2006 Unbreakable Linux plan, which saw it undercut Red Hat with enterprise Linux support. Despite the competitive threat, Mickos is unmoved. "I hope they do that," he said, noting that it would be seen as an endorsement of the open source database.""

Re:Oracle Sucks

[MyLongNickName]

Serious question: What exactly is the advantage of Oracle over SQL Server? I asked that to an Oracle DBA once, and he just got red in the face and stammered about having more options to configure things the way he wanted. I asked what exactly he configured, and basically got a lecture on Microsoft being evil. I then asked if he thought Larry Ellison was a saint, and the conversation just continued to devolve.

Serious question: why is Oracle considered so much better that SQL Server?

interesting

[larry bagina]

postgresql is a more oracle-like (eg, plpgsql) and BSD licensed. Of course, postgres could cannibalize oracle sales; mysql is like sqlite, but with less features.

Confused.

[Anonymous Coward]

Given that MySQL and Oracle are on complete opposite ends of the scale with regard to their uses, what benefit does Oracle have in killing what is essentially a non-competitor?

Knowledgeable IT people presumably already know when to use Oracle and when to use something smaller like MS-Access. IT morons who think that single-user databases with less than a thousand records need to be in Oracle have already drank the kool-aid and will never change to another product. Oracle's only threat would appear to be the next generation of IT morons who realize that Oracle is not always the best solution to simple problems. But since they don't know any better, presumably they will still go with Oracle.

Re:Oracle Sucks

[TheSpoom]

I've taken two Oracle classes (could have gotten certified, but I really didn't care that much). Basically, unless you're running into the tens of thousands of users, there's not much difference. Now, as a DBA, there are some things I would prefer to do in Oracle rather than SQL Server (though I'd much rather do them in an open source DB like MySQL or PostgreSQL), but I don't think it's worth the (massively) added cost unless you really are running it in an enterprise situation.

Re:Mikos is right.

[jimbojw]

> the primary benefit for Oracle is that they'll be able to offer bundled support with people who already have Oracle support and want the convenience of dealing with one company for all their support needs

This is not a trivial point.

Brand recognition goes a long way when a company is shopping for support. Companies that already pay for Oracle support might be very willing to tack-on MySQL support from them, rather than to establish a new relationship with MySQL.

Also, since Oracle will be distributing the MySQL database themselves, the unsuspecting middle-manager might think that "MySQL" is just another Oracle offering, or component of their architecture. It's the old "bundled with" implies "created by" heuristic.

Re:Mikos isn't right.

[mpapet]

Oracle has one kind of customer, MySQL has another kind of customer.

Just a guess, but I'll go out on a limb and state that any hopes MySQL had in wooing really pricey billable hour customers is evaporating. Even if I'm wrong, the mood at MySQL has probably been a little less happy when they figured out Oracle was going after the top of the consulting/support dollars.

There's still *so* much they have to offer for businesses willing to pay. They just need to keep at it and understand that Oracle won't be the first company to do this to them. Microsoft will surely follow with some kind of crazy scheme. They have to at some point as their arrangement with Novell suggests they need to at least appear as if they have something like OSS to offer.

Re:Oracle Sucks

[helixcode123]

What exactly is the advantage of Oracle over SQL Server?

I'm rather fond of their Analytic Functions [orafaq.com], which allow for convenient queries against other table rows. For example, given a table of time-stamped log entries you can write a query to "Show me the time intervals between successive log entries."

I'm hoping these will show up in Postgresql soon.

Re:Oracle distributing MySQL?

[martenmickos]

Thanks everyone for the interesting comments. It's a good discussion and I think there have been some good points made. I see Oracle and MySQL serving two distinct markets, so in most cases we are not competitors.

One point of clarification: I never said that Oracle has threatened MySQL. (I think this may have been the writer's editorial comment.) Instead, I view it as a positive thing that Oracle distributes MySQL. I have often suggested that Oracle should distribute MySQL and I've made the same suggestion to Microsoft. Perhaps Oracle, Microsoft or IBM will provide support for MySQL and that could be good for open source in general. (And note that Red Hat, HP and others already do so.)

If people want to buy support for MySQL from Oracle, that's great. And of course, if they want to buy it from us, we are fine with that also.

-Marten Mickos, MySQL AB

Re:Oracle Sucks

[Anonymous Coward]

Not quite true. I supported a large customer (a national, no nationwide, building society in the UK), that has several million customers in the DB, and a couple of thousand users directly logged onto it, and several thousand remote users. (this is SQL Server BTW), and its fine, fast and responsive.

Now, the things we had to do to make this work are basically: use stored procedures.

Now I work with a large breakdown company that uses Oracle, and we've seen barely a difference. Oracle is just as performant, and just as scalable. Oracle support wasn't too helpful with a recent 100% CPU issue ("install all patches" they said, then they broke something else and their answer was "upgrade to 10g").

One big difference that you'll notice when working with both is locking. Oracle basically does row-level locking, SQLServer tries to optimise locks and escaltes them to lock pages then tables as you lock more and more rows. If you write your app without taking this into account, you could have problems with SQLServer.

One thing with Oracle - use an application server to limit connections, don't try to write a client-server app on it, each connection hogs too much memory and holds query cache latches for too long.

But of course, these 'issues' aren't problems if you're not just slapping any old code together to run against the DBs. As with everything, they all need some care and attention with how you're working with them.

I have also used Informix and DB/2 on an AS/400. For really, really big DBs - give me DB2 anytime. that really is unbreakable, and IBM will give you a 'free' AS/400 specially optimised to run it on :)

Re:Mikos is right.

[Anonymous Coward]

considering that mysql is a cheap sonofabitch database without primary keys, and would not even remotely threaten OracleSQL for banks and their customers altogether, i can only supose that this move is done to field out postgress which may pose a serious security threat in the next decade.

Re:Isn't Oracle's database supposed to be unbreaka

[martenmickos]

Rorschach1,

Indeed we view MySQL competing in different markets from the legacy closed source databases. We have focused on new applications, often web-based systems, ecommerce, reporting, analysis and so on, rather than traditional ERP applications. There are many features that DB2 and Oracle have of which they are very proud. And we are also proud *not* to have all of the complexity of those features. Our focus is not on features, but on reliability, ease of use and performance.

Charles Phillips of Oracle remarked at a conference I was speaking at that Oracle and MySQL are both in the transportation business, but Oracle is the 747 and MySQL is the Toyota. I think that is a very apt analogy. But if you prefer the M1A2 tank, so be it. :-) In either case, we get very good gas mileage by comparison!

-Marten Mickos, MySQL AB

And you still don't have much of an answer

[Evets]

Oracle performs better in enterprise environments, hands down. Oracle Clustering is more intelligently implemented than SQL Clustering. PL/SQL scripts are easier to debug than those in MS SQL.

OTOH, SQL Server is extremely simple to install and administer for low volume environments. DTS Provides a nice simple transport mechanism. Enterprise manager, while kludgy, is relatively intuitive.

For fine tuning, Oracle provides finer control - but that's not to say that SQL doesn't provide a lot of control over DB Tuning features.

Then there are the little things that crop up over experienced usage - like the first time you try to take a MS SQL backup from one machine to another and end up perplexed for an hour. Or when you're 6 gig backup file won't copy from one machine to another without 3rd party software (really a windows issue, not SQL Specific). Or when you discover that you can't replicate certain tables or columns, can't copy blobs using sql scripts, etc. Things like that.

A lot of applications treat the database as a storage engine and leave platform specific performance enhancements by the wayside in favor of database-agnostity. Because of this, MySQL is much closer to being a legitimate competitor than you would think. People talk about "ACID Compliance", but really most applications don't need ACID Compliance and just because you can't do something one particular way doesn't mean it can't be done.

Re:I hope they do it for PostgreSQL, too.

[EugeneK]

The translation of some rather intensive PL/pgSQL to PL/SQL was almost trivial, with a translator script I wrote in a day.

any chance you'd release that script as Free Software?

Benefits of combining Oracle and MySql

[managerialslime]

The thread here seems to question (a) the value of Oracle and (b) why this would be good for Oracle's customers. Here are my 0.0002 cents:

My current client is a large insurance company. (More than $7 billion dollars of policies a year underwritten by a staff of more than 1,200 people.)

We have lots of Oracle, SQL/Server, and MS/Access applications all over the place. The Oracle data is generally available to everyone. We have more than 50 analysts who use a combination of Hyperion (formerly Brio) and SAS to model the data.

Oracle's additional configuration options enable our Oracle servers to support phenomenally more people than comparable gear running Access or SQL/Server. In addition, we have really good SAN devices that are backed up every night.

However, the limited number of Oracle DBA's mean that users must wait (sometimes forever) to get their application written in, or ported to, Oracle.

Where the SQL/Plus Oracle code is controlled, documented, and fully SOX compliant, the same is not always true for the Access and SQL/Server code.

As a result, the individual departments are forced to use Access and SQL/Server. Their applications do not talk to each other. The data in those applications are "hidden" from people in other departments.

SOoooo...... users develop "personal" and "departmental" applications in Access and SQL/Server which we in IT find and port to Oracle when we can.

MySQL apps are generally easier to port to Oracle than SQL/Server or Access. (Never mind the application layer. That is a different discussion for a different time.) I would love to provide MySQL to departments across the board on servers that are supportable by corporate IT. Then users could build their apps on-the fly, expect support from my team on an ongoing basis, and faster conversions to Oracle in the long run.

Wouldn't that be sweet.

Re:Typical support call

[fabs64]

I once did a "developer usability study" for some uni class, was told to basically get on the net, and set up the computer to be a development environment for developing a small dynamic webpage with a db backend.
Having never done web dev before I went with the biggest target, MySQL backend with apache/php front, literally took me 5 minutes to download and install MySQL with its nice developer tools, and maybe another 10 minutes to have a reasonable understanding of how those tools work.

Now, I've never installed the Oracle server or Client at home, but I have had to install the Client on my work machine and others, and the process here is... arcane, you run some bat script off a network drive which spawns a process that's invisible for all intents and purposes except in the process list, and you know it's done when it dissapears. Takes about two hours, you then have to check a log file to see if it was successful or not.
Then you fire up their oracle product installation thingammy with all its horrid slowness, and try and figure out which (badly labelled) buttons you click on to make it download the developer client.

Now, I'm not sure if it's always like that, this is quite likely just a torture process thought up by someone in our ITS department, but that is my experience with installing anything oracle related.

Re:Mikos is right.

[head_dunce]

It worked for Red Hat, now they have big money invested in their systems
http://biz.yahoo.com/bw/070129/20070129005032.html ?.v=1 [yahoo.com]