←

→

Updated database landscape graphic

November 2nd, 2012 | | Data management

One of the most popular pieces I have produced since joining 451 is not a research report or presentation but the database landscape graphic that accompanied our NoSQL, NewSQL and Beyond report.

We’ve seen it crop up in other presentations and websites – sometimes even with attribution 😉

We actually updated the image to accompany our more recent report MySQL vs. NoSQL and NewSQL: 2011-2015 but I realised that I haven’t made that newer version more generally available. So here it is:

We wouldn’t claim it to be perfect. There’s a whole new breed of data platform-as-a-service providers that have emerged in recent months that will need to be added, if we can find space for them.

Meanwhile there are a group of database vendors that have also emerged that don’t easily fit into the segments we’ve created: companies like Drawn to Scale, FoundationDB, Aerospike and Splice Machine.

But since the original graphic continues to be popular, I thought I’d share the latest iteration as well. Any feedback always welcome

Tags: , ,

23 comments ↓

#1 Nick Lavezzo on 11.02.12 at 11:34 am

Thanks for the mention, Matt.

It does look like it would be hard to find a spot for FoundationDB. What’s difficult on the chart is differentiating systems with and without ACID transactions. It used to be true that transactions was largely the domain of relational systems, but with the Google Spanner paper and the new entrants to the market (FoundationDB included) that is no longer the case.

#2 Arthur on 11.03.12 at 5:26 pm

Informix is not on the map at all?

#3 Matthew Aslett on 11.04.12 at 4:55 am

Good point. Slipped my mind

#4 Christoph Bussler on 11.04.12 at 10:09 am

Hi, a somewhat complete list of relational databases is here in graphical form: http://www.hpi.uni-potsdam.de/naumann/projekte/rdbms_genealogy.html

#5 Lilly on 11.05.12 at 10:09 am

Where would CartODB fit here?

#6 Matthew Aslett on 11.05.12 at 5:28 pm

Probably on a map application development landscape graphic

#7 adrian on 11.06.12 at 3:23 am

What about SQL Server? Sesame? Virtuoso? Big OWLIM? And other triple/quadstores?

#8 Matthew Aslett on 11.06.12 at 8:08 am

SQL Server is in there. I have seen no interest in the likes of Sesame or Virtuoso from our client base or wider circle . Room could and arguably should be made for the triple/quad stores in a future rev, although again I am seeing limited demand.

#9 MIke on 11.14.12 at 1:01 am

rethinkDB is missing?

#10 Matthew Aslett on 11.14.12 at 2:32 am

RethinkDB isn’t missing, although it was in the wrong place given its recent re-emergence as a document store. I have updated the chart accordingly

#11 Peter Hounsome on 11.14.12 at 7:52 am

Love it and may use it in a report – ‘We’ve seen it crop up in other presentations and websites – sometimes even with attribution” – wouldn’t dream of not doing it – any special requests for references/attribution?

#12 Matthew Aslett on 11.14.12 at 8:28 am

Thanks Peter,

Just a mention of my name, 451 Research and a link to this blog post is great.

#13 Dan on 11.16.12 at 2:37 pm

My how far they have fallen. You have left out Ingres and Sybase. The father and grandfather of SQL Server and PostgreSQL. Both products are still available. I think Sybase has a larger market share then Informix. Still this isn’t saying very much.

#14 Matthew Aslett on 11.19.12 at 3:52 am

Hi Dan,

Both SAP Sybase and Actian Ingres are on there and always have been.

Matt

#15 Baxter on 11.21.12 at 11:09 am

Well done! Of course there will be folks who ask “why not XXX?!?!?!” but I think you did a great job considering the multitude of options out there. Great to see you correctly pin Couchbase as straddling key-value and document.

#16 Adrian Otto on 12.15.12 at 11:00 am

The Rackspace “Relational… as-a-service” product is named:

Cloud Databases

Not:

MySQL Cloud

Please refer to it as “Rackspace Cloud Databases”

Thanks!

#17 Matthew Aslett on 12.17.12 at 4:18 am

Thanks Adrian, I had forgotten to update that. Now fixed

#18 mariuz on 12.23.12 at 1:59 am

You forgot Firebird in the Relational/SQL space , One of the best MySQL/Oracle alternatives

http://firebirdsql.org/

#19 Juign on 01.20.13 at 10:40 am

I think you skipped LDB – the in-memory Hadoop (http://highscalability.com/blog/2012/11/26/bigdata-using-erlang-c-and-lisp-to-fight-the-tsunami-of-mobi.html)

#20 Lance Norskog on 01.20.13 at 6:00 pm

Solr has morphed into a “Big Data” system with automated sharding. It is an open-source text search database in the Apache Foundation. It is not relational. In this picture it would be “key-value plus text search”.

#21 Peter Baumann on 01.28.13 at 10:06 pm

…and then there is rasdaman, an Array DBMS (not SQL, but in the spirit of SQL) supporting massive multi-dimensional arrays like satellite images, 3D image timeseries, 4D climate data – see also http://en.wikipedia.org/wiki/Array_DBMS .

#22 Dongya Cao on 03.01.13 at 2:49 am

Why PostgreSQL\MySQL\MariaDB are all analytic, but EnterpriseDB based on PostgreSQL is not?
Thanks

#23 Stefan Manegold on 04.09.13 at 5:56 am

Dear Matthew (et al.),

Thank you very much for putting together this very nice overview!

For completeness, I’d like to point you to MonetDB (http://www.monetdb.org/), an open-source relational analytical DBMS and one of the column-store pioneers. The current “production” version supports SQL. There is a version that supports XML/XQuery (http://www.monetdb.org/XQuery). Support for Arrays (SciQL: http://www.scilens.org/Resources/SciQL), JSON/JAQL (http://www.monetdb.org/Documentation/Extensions/JAQL), data stream processing (http://www.monetdb.org/Documentation/Extensions/Streams) and RDF/SPARQL is currently being developed.

Best,
Stefan