The spaghettified DBMS chart that shows Oracle's crown is slowly slipping

It might look like a map of the London Underground designed by a madman, but Gartner's newly-completed DBMS Market Share Ranks: 2011-2025 has an important message. The change may be glacial, but (most of the) dominant database vendors are slowly losing their grip on the market. Gartner database market share chart, supplied by …

[Anonymous Coward]

Scary. MS SQL is the worst of the lot.

[Anonymous Coward]

Yeah but outside of AWS it's the one that comes with the highest number of meetings and highest density of corporate bullshit...which is important when it comes to databases because it needs to be a project with a 6 month plan with an out of the box team based focused for the benefit of your customers synergising your partnerships to increase sales funnel throughput blue sky put a pin in it pie chart bollocks, not just somewhere you store data that you need to access from software with the highest performance and lowest overhead.

[Anonymous Coward]

The popularity of a database platform seems to be directly proportional to the quality of paper the brochure was printed on and how ironed the chinos are on the guy in the photo on the box.

[Anonymous Coward]

Strangely,

ms-sql, since they bought it rather than write it, and it was originally written for Unix, actually isn't as bad as you'd expect.

Don't run it on windows though. It assumes all memory is it's to use.

[Bebu sa Ware]

Re: Strangely

"ms-sql, since they bought it rather than write it, and it was originally written for Unix, actually isn't as bad as you'd expect."

Sybase wasn't it ?

A good few years ago I recall MS made ms-sql available for Linux (at least RHEL) - it might have been an evaluation license or developer/hobbiest license but I certainly didn't provide a CC or pay cash. I only installed it on the off chance someone had a need for ms-sql on RHEL. No one ever did.

A curious lifecycle Unix-Windows-Linux but not as uncommon as might be thought even within MS.

[JLV]

Re: Strangely

It was Sybase and when I was getting started on RDBMS in the early 90s, it was far ahead of Oracle (6?) with things like stored procedures and triggers.

Oracle marketing delivered a master class in risk containment however and successfully negated that advantage in a number of shops I saw, as management was always quick to point that Oracle 7 (not yet available and which did not come out till about a year later) did have stored procedures and triggers. Even then early Oracle procedures weren't as flexible as they lacked support for SELECTing data back to the client code.

(whether stored procedures and triggers are necessarily a good idea is another question, to me they're a tool in the toolbox, to be used or not).

[Triffids]

Re: Strangely

Oracle released version 7 in 1992, and PL/SQL was a proper language with packages and dependency checks. Sybase/MSSQL, even today, is just an SQL extension. As far as I know, even today in MSSQL, you'll get almost all errors at runtime.

[tiggity]

Re: Strangely

It runs surprisingly well in a container on Linux (never played with a Linux specific version of it so it may be dire compared to that)

[TKW]

Yeah. And it essentially contains its own OS / hardware abstraction layer which 1) is interesting 2) made the port [back?] to Linux possible.

[sindrelspec]

"Don't run it on windows though. It assumes all memory is it's to use."

Only if you don't configure it properly.

[Steve K]

That's only the case if you accept the default configuration on install

Absolutely.. That's only the case if you accept the default configuration on install

[Anonymous Coward]

Sounds like somebody hasn't used both MS SQL and Oracle, because Oracle is both worse and far more expensive.

When you see the name Oracle on a high profile project you know that it's going to be a high-profile fiasco that burns through cash in a way happens with no other vendor. There are reasons for that.

[Anonymous Coward]

Sure T-SQL was far better than PL/SQL... I remember working with MS SQL 6.5 and it could not store dates before 1600 (I was working on an application for a museum that had artifacts much older, with precise dates, often....)

As soon as someone uses a RDBMS a real database and not a simple data bucket the real differences appear.

[prandeamus]

Be fair, if you're storing dates for historical artifacts that can require a lot of functionality that isn't work including in generic functionality, like the Julian/Gregorian switch (Old Style vs New Style dating) which happened over the course of several centuries at different times in different countries. Or pre-Julian dates in Ancient Rome, or different conventions about the start of the new year. To say nothing of non-Western calendars. So yes, I dare say you had problems with MSSQL 6.5 but it seems an odd hill to die on.

It's great to have a turing-complete programming system on your RDBMS sometimes. PL/SQL or T-SQL or that limited .NET runtime thing. Having said that, if you rely on it too deeply for business logic other than simple triggers it can be a right royal PITA to manage.

YMMV

[jammy]

A little confused, as the article mentions suppliers and how much money they are making, rather than products. As such, Microsoft makes money from a number of database products, including MS SQL, Azure SQL, PostgreSQL, Cosmos, etc.

The same applies to AWS, which leads the table

[Anonymous Coward]

That chart though.

Jesus christ.

[Anonymous Coward]

Re: That chart though.

https://ibb.co/s9Zgt5Mg

[Bebu sa Ware]

Re: That chart though.

The Gartner coloured pencil department on overdrive - clearly an ADHD employer of choice.

A twenty channel chart recorder of a billionaire tech bro going full-rant ?

[MyffyW]

Re: That chart though.

That chart gives the impression of telling you far more than is actually conveyed. Which is very typically Gartner. I'm on the Crayola-spectrum, but that chart is - to wilfully misuse a technical term - triggering

[Groo The Wanderer - A Canuck]

By PostgreSQL thou shalt live and by PostgreSQL thou shalt die. Death to all who oppose PostgreSQL... :)

[Anonymous Coward]

That's a bit harsh. A week's worth of the medication taken prior to a colonoscopy ought to do it.

They'll be very quiet then. Lighter too, and they won't dare cough..

/s

[Paul Herber]

Cockroach Labs and Aerospike, start at the bottom and get worse!

[Anonymous Coward]

> The only dramatic change this year is the continuing meteoric rise of Cockroach Labs, the company behind CockroachDB

Surprised to see that the business model of "I'm going to copy (badly) a solution used by Google / Facebook to manage their humongous data and sell it to corporates even though *their* data is orders of magnitude less and could probably be handled with little more than a raspberry and a USB drive" is still alive.

> the RDBMS with a PostgreSQL-like front end and a distributed back end.

Translation for DBAs: it uses the PostgreSQL wire protocol.

[Fido]

The order is based on revenue of commercial offerings, not based on market share or job opportunities. So the ordering reflects something more useful to investors rather than developers.

[Richard 12]

Not even that, because it doesn't tell you anything about the gaps and thus very little about the direction of travel.

Are Oracle within a couple of percentage points of the one above - or the one below? Are they rising towards the one above or falling away from it?

This kind of chart is light on information.

[williamyf]

Is is part of the "Free-Lite" info package to drum up interest in the paid for dossier with much more information and cost a pretty £€¢¥$ ... Some of my employers of yore had those reports available for me to read, and belive you me, they went into excruciating detail.

Wether you trust gartner's methodology, data, or conclusions, that's another thing entirely, but do not criticize the Freebie for lack of info, that info is reserved for the Paid for report.

[Pascal Monett]

Re: This kind of chart is light on information

Well, it's Gartner. What did you expect ?

Looks impressive, though.

[sarusa]

The Heineken Factor

It's sad seeing Microsoft so high up there - but cheap terrible shite is popular in general, so why would it be any different here?

[williamyf]

Re: The Heineken Factor

Rememebr that, just like Oracle and IBM, Microsoft's entry in the chart is composed of standalone MS-SQL and The cloudy database services they offer on azure, than may include things like Postgress, mysql, or even oracle instances.

Also remember that of the big three standalone RDMSs (Oracle,IBM's DB2 and MSSQL) that can offer you support contract that even a DJIA company can accept, with all the bells and whistles (and throats to choke), microsoft SQL is by far the cheapest to buy both the base license and the support contracts.

Also, MS-SQL runs on linux quite well.

[MyffyW]

Re: The Heineken Factor

No love lost for IBM (or MS) here, but I'd cheerfully run either if the alternative was Oracle. Their sharp practices on licensing are truly dreadful.

[Gene Cash]

This ain't DBMS vendors, it's cloud

C'mon guys, can ya at least get the title right?

[captain veg]

Re: This ain't DBMS vendors, it's cloud

Yes. As such it entirely fails to mention exactly which databases are running on AWS in first place.

In my workplace we've been on SQL Server for donkeys', moved over time off-prem into a datacentre then an EC2 instance and latterly RDS. Nevertheless, I expect that most Amazon DB customers are running some kind of PostgreSQL.

-A.

[werdsmith]

Re: This ain't DBMS vendors, it's cloud

Well indeed, SQL Server on RDS is chucking money away. BYO licence on EC2 also requires Software Assurance for the licence mobility.

As good as SQL Server is, if your application offers the open source or AWS native option then that’s the way you should go.

[Cris E]

Re: This ain't DBMS vendors, it's cloud

Usually, sure. But corporate life is rife with shrink-wrap packages written for MS SQL or ORA or whatever and you just load them up because it's 100x harder to try moving them to some secondary platform (if it's possible at all.)

[David Newall]

PostgreSQL not included

Tell me they aren't measuring RDBMS without saying they're not measuring RDBMS.

Gartner must have a hidden agenda. But I repeat myself.

[LateAgain]

Re: PostgreSQL not included

Not including PostgreSQL just shows that it's only "commercial" vendors.

Especially since PostgreSQL can actually run oracle code, apparently.

[DrXym]

Re: PostgreSQL not included

Postgres has a PL/pgSQL extension that is similar to PL/SQL in Oracle. But it's not quite the same. Postgres also has excellent support for SQL, almost SQL:2023 compliant and so does Oracle, but both have their own proprietary types and extensions that are not portable.

So basically the best approach regardless of database is code to the standard wherever possible and avoid stored procedures unless there is a good reason you need them. Contrary to what you might expect, stored procs can actually be worse for performance than just a prepared statement. But if you do use them, they are close enough to be ported over.

[doublelayer]

Re: PostgreSQL not included

Tell me you didn't read the part of the article that explained that without... It's measuring money paid to a specific company. Since I didn't pay anybody to deploy my latest PostgreSQL instance, they can't measure it on this chart. As others have pointed out, that makes the chart they were able to make not very useful, but it doesn't mean they were intentionally trying to leave out someone to promote an agenda.

[williamyf]

Re: PostgreSQL not included

Not very usefull for you as a developer.

But for some analyst who is running a diversified tech portfolio in an investment firm, and wants to have dome exposure to RDMS, is very usefull data. Especially since that guy/gal will not get the freebie, s/he will get the full paid for long form report.

After all, if you wanted to invest on "mySQL" ¿Which companie's share would you get?.

Yeah, tought so.

[doublelayer]

Re: PostgreSQL not included

I suggest it's not the most useful for them either, since if they can't measure actual usage, it's harder to project the long-term performance of the vendors. This report mixes spending on cloud and spending on software, so if you're trying to limit yourself to databases alone, you have to do more work to figure out how to do that without polluting your data with infrastructure companies who are likely in a different sector of your diversified tech portfolio.

[Anonymous Coward]

Re: PostgreSQL not included

> After all, if you wanted to invest on "mySQL" ¿Which companie's share would you get?.

Oracle's.

Or do you mean MariaDB, or something else?

In that case, I would:

* Be actually knowledgeable about databases

* Be knowledgeable about the markets and context in which the product (say, MariaDB) is being used

* Look at their Git repository to see who is doing the heavy lifting and what are his employers / sources of finance

* Look at forums, etc, to see who is writing (in an informed manner) about it and answering users questions and research the same points as above

What I would *not* do is pay £2-3k for a stupid report full of fancy charts written by people with shallow or no expertise in the subject domain. I've made that mistake in the past.

[LateAgain]

Is it bad that I misread the headline?

Oracle's Clown.

[trevorde]

Oracle

Just say 'No'

[trevorde]

Re: Oracle

Wow! A down vote... Is that you, Larry?

[trevorde]

ABO

Anything But Oracle

[Groo The Wanderer - A Canuck]

Re: ABO

I can't agree. I'd rather use Oracle than Microsoft SQL Server. Bastardized Sybase 10 that is - obsolete technology from the ground up.

[Bebu sa Ware]

Re: ABO

"Anything But Oracle"

Unfortunately that also encompasses oceans of misery around the Oracle Sargasso.

I shudder to think of implementing anything today using Unify RDBMS (v4.2) from the 1980-90s and it wasn't too bad at the time.

[phuzz]

Re: ABO

I did appreciate how delicately the article managed to avoid directly saying the obvious: Nobody chooses to use Oracle, but many companies are stuck with them because migrating would cost even more.

[Cris E]

Re: ABO

Ah, the Inertia Tax weighs heavy.

[Anonymous Coward]

Far fewer reason to use Oracle today

Once, on the available hardware, there were few options for many medium/large database workloads. Oracle could scale and could run on hardware that could allow that scaling, while offering required capabilities for point-in-time restore and disaster recovery.

There was a time were the low-cost alternatives were the dBase clones or Access (aaargh!)

Today for many workloads that advantage went away, and Oracle costs and commercial policies became just an unjustified burden. Many workloads con run on MySQL or Posgres - while other works better with new kind of databases like Mongo or Elastic.

Oracle got MySQL while acquiring Sun to control Java, but wasn't able to change its business model to adapt to the new landscape. Just like Broadcom, they believe they can make enough money from big enough customers alone, but they too could find they can be squeezed up to a point.

[Anonymous Coward]

Re: Far fewer reason to use Oracle today

> while other works better with new kind of databases like Mongo or Elastic

I still have to find that "other" that turns snake oil, whether branded "Mongo" or "Elastic" into something useful.

[JLV]

The Postgres treatment in this article would have been improved by noting that it has some revenue-facing manifestations: EnterpriseDB and some vendors' db-as-a-service offerings with things like AWS Aurora that do figure in the Gartner chart.

[DrXym]

And AWS even rents specialised EC2 instances for Postgres

[anonymous boring coward]

The chart only shows ranking, not actual numbers. So small changes can look like massive ones. Must have been an MBA designing it.

[Anonymous Coward]

Rankings can be deceptive

Ranking by revenue but not scaling by it compresses the chart - I'd like to see the same chart with real scales as I suspect there would be quite a large gap between the top 5 and the middle ranks that is not visible from this report, and gives a skewed impression of some companies' relevance / standing...

[Berny Stapleton]

The old Microsoft backoffice logo

Is it just me or does the top three lines in the 2019 - 2022 range remind anyone of the old backoffice logo?

[Sproggit]

Don't Forget the Q

Having read the article and the comments, my sense is that both are discussing the ebb-and-flow of database views over the last 20+ years or so. Go back to the late Nineties/early Noughties and it was Oracle-on-Solaris as far as the eye could see. That was the combination to beat because Solaris ran Oracle better than anything else - and Oracle had the speed, scale and flexibility to run anywhere. What's happened since has been a steady drip, drip, drip of evolution.

But all of this ignores the fact that we're starting to see some really promising developments at the intersect between Quantum Computing and database design. Quantum computing is bringing features and capabilities - like superposition, entanglement and quantum interference - that can directly impact routine, common database operations.

Grover's algorithm offers a quadratic speed increase for searching un-sorted databases - which can reduce search times from minutes to milliseconds for massive datasets - which are becoming more prevalent thanks to AI and Cloud...

The development of Quantum Partitioned Database solutions [QPD] are opening the doors to more efficient data insertion, deletion and selection - and the more complex the data model, the greater the advantage it provides...

With transaction management - methods such as Quantum Walks can massively improve transaction scheduling conflict resolution and throughput.

Basically, quantum computing brings parallelism, reduces the need for steps like pre-sorting, and improves vector search or the definition/annotation of complex data patterns with much greater speed and reliability than any conventional technology.

Right now the big limitations are NISQ related - the qubit count of available quantum computers, decoherence and of course more mundane things like the cost...

But pretty much every technology mankind has perfected started out with the same challenges... and we overcame them all.

I don't think "Oracle's Crown" is the right question to be asking or the right observation to make. I think that if you want a clearer insight in to the state of [top end] database deployments in the mid-future onwards, you need to look to what these companies are doing to marry their database tech with quantum implementations; who they have working those problems; what their use cases or drivers are.

As the technology is brought more in to the mainstream, it will be those with the insight that move forward the quickest.

[Anonymous Coward]

May be interesting to look on the graph at Software AG.

DB software designed for big iron (& destroyed Oracle (when they still aimed at mainframe), DB2 etc in performance terms)

Impressive in their niche but they never really managed to transfer to other platforms, so the graph makes their decline clear.

A good illustration that "dominance" now does not equate to future dominance in a changing landscape & might give some Oracle shareholders the occasional sleepless night *.

Anon as former SAG employee

* Though probably not as much as Oracle going all in on "AI" might!

[Anonymous Coward]

Oracle has another major problem on top of absurd price tag: Ellison is a major asshole as a person.

Not only that, he uses corporate money to fund Trump and his horde of idiots.

If you don't want to do that, you do not pay for Oracle.