Oli Sennhauser is the CEO of FromDual, an European consulting company specializing in the MySQL database and its derivatives like MariaDB or Drizzle. Earlier this year he joined the Open Database Alliance as a Service Provider member.
In this interview Henrik Ingo takes his pulse on the European open source database landscape.
Rather than avoiding the awkward question, let me just get it out of the way first: You left Sun Microsystems earlier this year and started your own business. What motivated you? Was this because of the Oracle acquisition?
Not only. The Oracle Acquisition just was the straw that broke the camel’s back. Since years I tinker with the idea of running a company on my own. But always a job which was interesting enough to hinder me crossed my way.
Then as a consultant within MySQL I found the ideal situation to do technically challenging work with open source technology on a high pace with interesting and innovative customers. This was a very satisfying job. I fear that this customer segment is not so interesting for Oracle and they will disregard it. So I decided that NOW is the right time to start my own consulting company.
What kind of services is FromDual selling? Do you have some special niche or strengths?
We are focussing on vendor independent and neutral MySQL/MariaDB consulting and customer tailored training. This means we can show the customer where those products are the right choice and where not. We also can show the customers where are the rough edges of the products, where he has to be more careful and where the products are solid. That is something you cannot expect from a product vendor. Because we have not our own Database product or Storage Engine we are not biased and our only goal is to find the best solution for the customers needs. There is already a broad choice of MySQL derivatives (MySQL, MariaDB, Drizzle), Storage Engines (Maria, XtraDB, PBXT, etc.) and related technologies (Replication, Clustering, etc.) so customers not too familiar with the matter can become really confused. At the moment we are the only independent MySQL consulting company which can provide German speaking consultants. We strive to keep our consulting engagements in the customers local languages. At the moment we can cover German, English and French.
In the USA I know of many independent consultants in this space, starting from one man companies to several with 20-30 employees. But in Europe you seem to be the first one really specializing in the MySQL layer of the stack? What is the European MySQL landscape like nowadays? I know many PHP or Linux consultants also can support MySQL, do you compete (or collaborate) with them?
Yes indeed. It seems like everybody is/was concentrating on the US market. But here over in Europe we have a huge amount of SME’s which are using MySQL. Because everybody goes West, I decided to stay here and help the MySQL users here in Europe. Also the customers in Europe need someone who takes care of them.
And most of the companies here in Europe prefer to have somebody on-site in their local time zone and preferably in their local language. These are different conditions and possibly a bit a different mentality as in US.
Yes. I cannot imagine I would just SSH into the servers of some of my past European MySQL customers, which is apparently common for, let’s say, Percona or Open Query customers.
… How the landscape looks is a difficult question. And it depends heavily from which point of view you look at it. We get customers from a little 2-man company which need help recovering their CMS database back-end after a crash over middle size companies which do really fancy stuff with their products and MySQL as a back-end up to international world-wide operating companies which want to safe cost with MySQL or use the hyper-fast and highly available MySQL Cluster.
And finally, you are right. Many developers and administrators have some know-how about MySQL. Databases seems to be not sexy enough to concentrate on them. They remain always the stepchildren. As a consequence most of the customers or the “generic” consultants can solve the simple things but not more. FromDual consultants come into place where it becomes tricky. We try to solve the problems where you do not find the answer on every second MySQL related web-site or where the problems are too complex to answer in a few sentences or to solve it in 30 minutes work. So as a result we somehow collaborate with those guys and do not compete with them.
So tell me, how are you getting started? You just completed a long consulting gig, so it seems there is a lot of activity and demand out there?
Oh, this engagement is not over yet. We are just finishing phase one of the project. After summer vacation my customer wants to continue with phase two and it seems like our involvement is needed again. But basically you are right. There is plenty of work to do out there. The MySQL/Maria ecosystem is vibrant. Also over here in Europe. A lot of activity means a lot of friction and heat so we have to cool down the situations at customers and make their databases work again fast and smooth.
Already from the very first day there was something to do. I stopped working at my former employer on Monday and already in the same week I had my first engagement at my own customer. I have informed my network and it helps me looking around for possible engagements but most of the customers up to now came by their own. They had a problem, where searching around, somehow found our company and finally became our customers.
That’s excellent to hear. It takes a bit of courage to start your own company, so having work to do from day one sounds really encouraging! And it’s good to hear you are filling a need.What kind of things do you see European MySQL users doing? What kind of projects are you seeing?
We do a continuous MySQL ecosystem research. So those are MySQL users and not necessarily customers. What we can see there is that the clear number one usage of MySQL is as a database back-end for any kind of CMS (mostly Typo3). Then a lot of companies use MySQL as database back-end for any kind of company home grown application. And then there are some companies which use MySQL as database included in products or services which they sell and ship to their customers.
Projects we are seeing at the moment are mostly from this second group. For example one company does monitoring of radio and TV consumption. After some really fancy technical gimmicks the data end up in a MySQL database where they generate the daily reports. So if you do advertisements over TV or radio you know pretty well how many people have listened/seen your advertisement. Another customer has its whole production process based on OpenOffice spreadsheets. This is not satisfying, not real time and does not scale at all when they will ramp up their production by factor ten and spread it over different locations.
Ah, I remember working for a training company that would run its entire curriculum, with dozens of trainers, on a set of Excel spreadsheets. But that’s not all, when I started with databases, I have literally migrated a box of physical paper cards into a SQL database. That was in 1996 :-)
A big database vendor which now also offers hardware made them an offer for a 6-digit Euro solution. We showed this customer a solution which perfectly fits his skills and needs for a 4-digit amount of money. The solution will scale most probably until he makes several hundred millions revenue. If they reach this point we will see further what his needs are. And all is based on open source technology. So he has no vendor lock in, he is free in his choice, has everything under his own control, and will stay flexible also in the future.
Those are just 2 examples of interesting projects we are seeing at FromDual.
Thanks for sharing that! That kind of takes me to the next question I’ve been dying to ask:
A personal interest of mine – a “mission” of sorts – was always to encourage those enterprise users who mostly run on proprietary databases to make the switch to an open source database. I don’t see MySQL/MariaDB being any different than Linux is for operating systems in this regard. Do you see MySQL being considered for that, or have database customers become concerned about MySQL’s fate being uncertain?
A very interesting point! What we can see, the situation is two-fold. One one side there are customers who trust in big Oracle and thus in the future of the MySQL database. Here we can see efforts of those customers moving away from the Oracle database product itself to the MySQL database (which belongs to Oracle now). Those customers we can help with a smooth migration and help to build their new system properly.
Then there are customer who do NOT trust in Oracle. Those customers are looking around for alternatives. Some of them consider PostgreSQL as an alternative but a lot of (Open Source) application nowadays mainly run on MySQL and are not ported or fully supported by PostgreSQL. When we show them the possibility to go for MariaDB which is an drop-in replacement for MySQL they are relieved and very happy that we showed them this back door. This is the advantage of a vendor independent and neutral consulting company as FromDual is. :-)
Interesting. I guess that too confirms what I’ve been kind of seeing. I guess it is no secret, I’m myself is one of those in the group with some distrust. At the same time I can see the big interest from Oracle users, such as reported by Giuseppe Maxia from the meeting with International Oracle Users Group or Ronald Bradford and Sheeri Cabral organizing the MySQL track for the Kaleidoscope Conference of ODTUG. There are many Oracle users out there who now have an interest, or maybe even “permission” if they have restriction on which database suppliers they can use in the company, to look at MySQL for the first time.
There are now quite many forks of MySQL, and separately of InnoDB too. What do you think about this situation and where do you see it going?
What you can see in markets ever and ever again is at a certain point a diversification process. This has started with MySQL in 2008 with the ProvenScaling branch. If this diversification process has already reached its maximum I do not know but I believe so because there is not too much folks out there which is capable to fork or branch MySQL properly and know what they are doing. And all those who know have already done so. After the diversification the consolidation process takes place. And for this step we can see already the first signs. As far as I understood all the different branches are now merging into the MariaDB tree. So it looks like MariaDB is sooner or later the most feature rich MySQL derivate and could become the new landmark if it is not already. But how far it has progressed you probably know better than me.
At the moment it looks to me like there will remain 3 branches/forks competing for the leadership after the consolidation has finished: MariaDB, Drizzle and MySQL.
If Oracle can follow the high pace and motivation you guys are showing we will see.
From the storage engine point of view it is also very interesting because here we have less restrictions and a much more “free market” which means more competition. At the moment it looks like XtraDB is a bit ahead of InnoDB but this can change from release to release and it depends on which Benchmark Buzz you believe and what you benchmark exactly. But from the point of view of most customers it does not matter if you choose XtraDB or InnoDB. Now there is a new very interesting competitor in the market: PBXT from Paul McCullagh which shows promising results. It seems like in some situations it is even faster than InnoDB/XtraDB and it works on MySQL/MariaDB and Drizzle. If nobody plays a dirty game by complicating the use of the Storage Engine API it will be an interesting future and possibly a benefit for the customers in the end.
And what is your plan with your company. Do you want to stay a one man show or do you have something bigger in your mind? And what about cooperation with other companies?
I have seen from my previous employments in different consulting companies and by visiting customers that a too rapid growth is not healthy for a company, nor healthy for customers, nor the employees.
So my plan is to grow the company moderately. If everything works according to plan we will hire one or two more MySQL consultants in the next few months. Then consolidate and then see further…
If we have a big enough core staff we can think about offering new and more sophisticated services to our customers. For example we could run their whole MySQL/MariaDB database infrastructure. So they do not have to hire their own expensive staff.
Currently we are working already sporadically together with other consulting companies and we are a member of the ODBA. So we help out each other if we lack resources or know-how. Or we pass over a deal we have not the capabilities to offer ourself because of lack of resources. So there is a loosely coupled network of companies around ODBA spread over the planet which does its best to help the MySQL/MariaDB community.