Ivan's Blog

Still blogging after all these years

2011-12-09T23:05:00.001Z

Or, better say, I have not blogged for so long…

What happened? Well, it has been a crazy year, probably the busiest year in my whole working life.

When you are in your late twenties/early thirties, you can dedicate all your energy to something, work, hobby, charity, sport, study, you name it. You work hard, and every now and then you have time to lie down and recharge your batteries. In one or two days, if you can sleep 9/10 hours per day, all the tiredness you have accumulated goes away. I have felt that feeling, but in your late forties, there is no such thing anymore.

When you are forty and you have a family, your priorities change. You try to accommodate work, family and friends. You remove most, if not all, of the time you dedicate to yourself. You have less rest, usually not enough to recharge. The 9/10 hours you used to recharge and be up and running, ready for a new week, they become barely 6/7 hours, if you are lucky… and you feel like your head is in a constant foggy atmosphere that does not allow you to think straight. You do, but you must use more energy and it takes more time simply to get things done.

So, this is what happened to me. But I was used to that. Even in the weird days of the post-acquisition, when many people could not really do much business-wise, I felt pretty busy.

Nevertheless, I managed to enjoy this fantastic year at SkySQL. 2011 has superseded 2005 (the year I joined MySQL AB), as the best year of my entire working life. Thanks to SkySQL, thanks to all my great colleagues, to all the partners, users and customers. And thanks to my family and friends, who are supporting me so much.

What are the plans for the future? In long term, it is hard to say, but the short term plan seems really exciting, more than ever, better than 2011. I can see so many new opportunities to use MySQL and MySQL-related products and to be involved in very cool projects. MySQL is not the only wild kid in the block anymore: PostgreSQL and the NoSQL armada are rightly claiming their space and users are adopting old and new technologies. This is good news, it repositions our technology right were it should be, leaving PostgreSQL, HBase, Cassandra and others doing what they do better.

Our technology - i.e. MySQL and the MySQL Ecosystem - is evolving and improving. There is a growing demand for database systems, there is space for everybody. There is space for new projects and for integrations with the current and new infrastructures.

SkySQL will be more and more focused on new solutions and on new architectures. The reference architecture is key for these new solutions. I will blog more about in in the near future and clarify the reference architecture is and what is not. We will offer legacy systems, we will support the standard, vanilla version of MySQL, the servers that Oracle and Percona can now provide. But we will offer more, much more. There is a whole new set of highly available systems - see my talks at various meetups in the US this week - there is a new way to integrate MySQL with other databases, and a new way to create huge data warehouses systems and compete head to head with Vertica or Greenplum, not to mention the quite outdated technologies in 11g. There is a huge demand of migration from legacy systems into the cloud, with MySQL.

Here is the promise - an End-Of-Year resolution, or an anticipated New Year resolution. I will share more of this experiences and the excitement of applying the open source values and technologies where until now only expensive commercial solutions were allowed. Some of my posts will be generic and will fit in here, in my personal blog. Other posts will be more MySQL specific and will be published in MySQL4all.

The future looks bright, I hope many will enjoy the ride with me.

-ivan

Finally: MySQL Meetup in London!

2011-03-20T22:23:00.000Z

Sponsored by Canonical and SkySQL, MySQL Users are going to meet again

The last meetup was nine months ago, in the "old" Sun Briefing Centre, the last event that was hosted in King William Street before Sun/Oracle closed it to concentrate the London offices and meeting rooms in Moorgate.

On Tuesday 22 of March, we will be guests of Canonical, in their prestigious headquarters at the 27th floor of the Millbank Tower in London. My guess is that the breathtaking views of London and the Thames will distract us for a while, but hopefully we will have enough time to go through a pretty packed agenda.

The preliminary contents, as usual, are subject to change at the very last minute. Everybody can submit a topic, in any format - presentation, white board or simple speech - and we will devote at least 10 minutes to it, or as long as the audience is interested in discussing. Andrew Poodle and Mark Baker will be the chairmen, I am sure they will do a fantastic job.

For the moment, we have requests around MySQL 5.5, MariaDB and NDB/Cluster. In true MySQL meetup spirit, we will head to the Morphet Arms to continue with more specific discussions and opinions to share.

Here are all the details:

This is the registration page: http://www.facebook.com/event.php?eid=134352579966946&ref=mf

When: Tuesday 22 March - 18.30 - 21.30

Where: Canonical, 27th Floor, Millbank Tower

21 - 24 Millbank

London, United Kingdom

Afterwards: Morphet Arms - SW1P 4RW

SkySQL Enterprise Products and Services: Welcome to 5.5!

2010-12-15T21:28:00.000Z

A bit early for Christmas, or perhaps slightly delayed for Saint Lucia, MySQL 5.5 is finally GA. A fantastic present from Oracle for the whole MySQL Community, something that many wish unwrap and install as soon as possible.

Tomorrow, Thursday 16 December @ 9AM PT/12PM ET/1800 CET, we will present the first webinar around the offering of products and services from SkySQL. If you are interested, you can register here: http://www.skysql.com/en/news-and-events/webinars

We have not blogged much about what SkySQL can offer to the Community and to its customers, so I think this is the occasion to understand how we operate, what is our position with regards to the main forks and distributions of MySQL (including 5.5) and which technologies we use and recommend. I think (and hope) the webinar will clarify these aspects.

If you cannot attend, don't worry, you can always download the recording later from here: http://www.skysql.com/en/news-and-events/webinars/on-demand-webinars

…but you will miss the opportunity to ask some specific questions.

See you online tomorrow!

-ivan

Taking off again, up high in the sky...

2010-11-11T00:23:00.000Z

After my departure from Oracle, I received lots of emails and messages, people who asked me where I will land. Thanks to everybody for the kind words and the great comments. By the way, more than a landing, I am going to take off… :)

On October 29th I joined Ulf Sandberg, Kaj Arno and Mick Carney at SkySQL. I will lead the Field Services team in Europe, i. e. my team will help users and customers in designing and implementing solutions around the MySQL technology. This will be my main role but, as you may expect, in small companies (small compared to large Corporates like IBM or Oracle), you must wear many hats.

My long term goals have not changed: I want to serve the MySQL Community of users, developers and, of course, the SkySQL customers. I want to provide the best possible advise, the best services and solutions. I will be specifically focused on these aspects:

Continue to support MySQL as the de facto standard for online applications

MySQL is great for online applications, but it is far from being perfect. MySQL 5.5 and MariaDB 5.2 show significant improvements in terms of performance and scalability, more statistics and admin features. We are still a bit behind schedule with real improvements, like the adoption of new storage engines - specifically PBXT and more NoSQL-like engines.

Scout and implement solutions that will make MySQL really suitable for the Enterprise and ubiquitous in new areas

In one of my old posts I have explained the difference between the mindset of an Enteprise-focused IT Manager versus an IT team devoted to the cause of open source. Part of the work to do is based on evangelisation, but there are some technical aspects to cover as well. MySQL needs some extra features to be "Entperprise-ready". "Enterprise features" are not just sophisticated features, or "all the features you may think of". They are features that are necessary to sustain Enterprise software and tools. The debate around which features are necessary is meaningless, since every Enterprise is different. There is a set of features that is quite mainstream. In particular, we need to improve:

- Documents and Multimedia objects handling
- Security
- Analytical functions and large data analysis
- Data sharing and integration

These improvements are not necessarily related to deep changes in the server core. Most of the can be implemented as plugins or at a connection level. Security plug-ins, Analytical storage engines, XML functions and connectors are still valid options that can well provide great features without affecting the core of the server.

Also, many tools for the Enterprise would simply facilitate the development and the administration of bespoken code or third party software. Tools like graphical set ups and monitoring for HA environments, replication or data migration, and so on would be incredibly helpful.

Promote and participate in new ideas that can make MySQL even better

If we have freedom to think, why not be open minded and think bigger? 1 out of 1,000 ideas may be a great idea! The amount of tools, engines, appliances, services, components and other products - commercial or open source - that have been developed in the last 3-4 years around MySQL is pretty amazing. Projects like Drizzle, storage engines like Infobright and Infinidb, appliances like Kickfire are great examples of these new products. Many others are incubated at the moment and they will bright soon. Some of these products can make MySQL a real success in many areas currently ruled by expensive and legacy solutions.

Make MySQL an affordable and inexpensive alternative to the other databases on the market

What is #1 reason to not use MySQL? Put it simply, it might not fit with the required solution. In the Enterprise market, it means that MySQL may not be certified to work with 3rd party products. In other areas it may be that other databases are the mainstream for certain projects. But IT solutions are really changing. Cloud computing is a reality and #1 database is, again, MySQL. SAAS solutions are the fastest growing products in IT and again, MySQL is becoming more and more the favourite database for a combination of robustness, reliability, performance, and obviously cost of ownership. At SkySQL, I believe I will have the opportunity to address these markets and to provide valid solutions for the Cloud and SAAS companies.

Maintain independence and objective views on the solutions available around MySQL

Despite the suspicion of some users, my ex team and I have always provided an objective view of the possible solutions around MySQL. When we were clearly not in the right place, we simply walked away, saving everybody's time. I want to maintain this level of independence and objectiveness and even more, at SkySQL I have the opportunity to experiment even more technologies and to be 100% agnostic, under any circumstance.

Contrary to what I posted few weeks ago, I will not close this blog. izoratti.blogspot.com is my personal blog and I am going to use it to post comments and ideas. The posts will be mainly related to MySQL, but sometimes I will diverge by sharing my personal experience on other topics. mysql4all.wordpress.com is the place where I will post technical evaluations and suggestions. I will get help from other people as well - the main objective is sharing useful information. I will also continue to work with the MySQL User Groups in UK and Italy, now more than before.

As usual, my postings will not be very frequent. I will let many other people write great articles and share valuable information on a regular basis. Every now and then, I will pop up on Planet.

See you soon!

-ivan

Good night, and good luck

2010-10-27T16:00:00.000+01:00

“We will not walk in fear of one another. We will not be driven into an age of unreason if we dig deep into our history and remember we are not descended from fearful men.”

-- Edward R. Murrow

When I thought about the content of this post, it came to my mind one of the movies where George Clooney is actor, director and writer. The movie is Good Night and Good Luck, a great snapshot of post-war American history and a great way, in my opinion, to analyse how a modern and democratic society has evolved since then.

On Thursday 30th of September, I resigned from Oracle. I left my role as Sales Engineering Manager in EMEA. I left my team, my colleagues and my friends but, as I said in a goodbye letter, I am taking with me all the friendship and the teaching of the 5 most beautiful years of my working life.

It would take too long to mention all the great moments I had at MySQL/Sun/Oracle - they are far way too many. And I would not name my colleagues either, it is fair to say that all my colleagues have been fantastic, with no exception. Kaj Arno recently posted a great message in his blog and I cannot agree more with his comments and analysis. MySQL, in all its forms, has been a fantastic place to work. And because it has been so fantastic, and in my opinion it still is, I found it is so difficult to leave. When one has experienced such a great time and passionate work, one hardly wants to move to something different. After many years involved in the sales of closed source products, I realised how exciting it is to work in a open source environment, I breathed this sense of openness and fairness. I embraced the culture and a more relaxed approach with users, customers and colleagues, and I did not want to go back to the days of closed source.

MySQL AB and the good old days days of the small company that owned MySQL, animated by the enthusiasm of its employees, customers and users, are over. It has been replaced by a far larger organisation, with more power and more resources, by many other company operating in the MySQL ecosystem and by a wider MySQL Community.

Oracle now has the ability and the resources to bring MySQL into new customers and new environments, in places where MySQL AB could not get in. This ability will serve the MySQL Community and users much better that the capacity of a single, small company. It will make MySQL even more ubiquitous, expanding in different customer-based. I can guess there are some skeptic readers frowning here. Let’s wait and see, time will tell.

If I am so optimistic about the future of MySQL at Oracle, why did I leave? I think that my personal attitude does not fit with large organisation. I believe that Oracle is a fantastic company to work for, but it is not the right place for me, at least not in the form and in the position I was.

What's next for me? I am clearing up and closing all my tasks at the moment. My mind was rumbling with ideas, some of them were crazy, some more conservative. Watch this space, it will be soon time for another post.

Good night, and good luck.

Moving to WordPress

2010-09-22T01:05:00.000+01:00

This is my last post on Blogger. I decided to move to WordPress and leave, at least for now, this blog.

I decided to move to WordPress because I found it more suitable for my needs.

So, from now on, you can read my posts re MySQL on mysql4all.wordpress.com.

My personal, non work related blog is izoratti.wordpress.com, but you may be disappointed, there is only a "Hello World!" post and I am planning to leave it like that for some time.

All the best to Blogger!

Ready for the User Conference?

2010-03-25T13:00:00.000Z

I cannot recall any significant moment of the conferences in 2002 and 2003 (simply because I was not there) but…

In 2005 we had MySQL 5. Peter Zaitsev was still working in the benchmark team for MySQL AB. His presentation on InnoDB performance and tuning was enlighting for many.

In 2006 we discovered the Pluggable Storage Engine API. Jim Starkey joined MySQL AB and we announced Falcon. [A brighter note,] I have been so lucky to meet Paul McCullagh the day before the Conference. Paul is one of the nicest and most brilliant persons I have ever met.

2007 was all around 5.1. We announced the roadmap for 6.0 and our online cross engine backup.

In 2008 we were Sun and for the first time Marten Mickos left his place on stage of the UC to Jonathan Schwartz.

In 2009 we had the Oracle announcement and the Percona Conference. You may describe the conference in many way, it certainly wasn't boring!

And now, 2010. Another User Conference with tremendous content. It's an incredible occasion to learn from the key players at MySQL and in the MySQL ecosystem.

I will present at the User Conference. For once, I am back to my roots, i.e. Data Warehousing and Business Intelligence. DW projects filled my working life from 1994 until 2005.

Do not expect any elegant fragment of C code that may improve the performance of your DB 100x (or it may crash all your servers). You'll see just real life ideas and solutions on how to use MySQL in Data Warehousing and in a typical (is there one really?) Business Intelligence environment. And perhaps you have already implemented something similar, or something that suits you better, but hey, sharing is the main point here.

This presentation at the UC is just the beginning of a series. More to come, since this is a hot topic and users are asking more and more from MySQL in this sector. They want to use MySQL in BI in many, many ways. Some are simply looking for a reporting platform - you replicate your data and there you, you execute some reports. Others have more specific needs and they must transform their data into information, in a typical BI style. Some others are Enterprises with large data warehouses and they see MySQL as the perfect data mart engine. These topics and more in my presentation at the UC.

If you are part of the local London MySQL group and you can't travel to CA, don't worry, we will a series of meetups and they are likely to be recorded and presented in other cities as well.

In any case, I hope to see you at the UC. Check the message board for some Euro spots!

Back to MySQL

2010-03-20T09:36:00.002Z

Yes, it's been a while. I did not blog for quite a long time. In April, I posted an article on MySQL on Solaris, and with that article I meant to go back to my regular blogging, i.e. writing an article every week or so. Soon after that article, I took responsibility of the European Presales team for Sun Software, Application Server Platform to be precise. APS included MySQL, the Identity Management software, SOA products, Glassfish and Java. This job was very similar to what I used to do between 1998 and 2002 and I was really excited to work with a large team again.

Now that we are part of Oracle, I am back to my original role MySQL, in the European Presales team, now called Sales Consulting.

It's really nice to be back. I really missed the excitement and the hands on that characterises the MySQL business. APS was a fantastic place to work and the incredibly talented people I've met are genuinely one of the best team I have ever worked with. But my soul is with databases and even more with OSS. I like the way we go to market, we approach users and customers and we design solutions. I like the debates and the passion behind important topics. In one sentence, I feel like I fit well and I am looking forward to enjoying this new ride with Oracle.

So, next call, the User Conference. I can't see a better way to start.

Glad to be back!

[Finally] Slides posted for the DW Breakfast in London

2010-03-08T17:49:00.000Z

It took a while, but here they are. We have posted only the slides from Sun/MySQL, since the other material is copyright by Infobright and Talend.

MySQL Data Warehouse Breakfast

2010-01-30T11:19:00.000Z

Do you think that MySQL is only for web applications? Well, think twice. MySQL is a pure relational database and with all the features pumped up in 5.0 and 5.1 it can be a perfect data warehouse machine.

Not convinced? Well, let me help you. Next week, on Thursday 4th of Feb to be precise, we will host a new breakfast event in London on this subject. Considering the limited space available at the venue, the event is by invitation only, but If you are around London and you are interested, you can drop me an email at ivan@mysql.com with all your details. If there are still seats available, I will be glad to grant you a pass. If you can't participate, don't worry. As usual, we will post slides and comments on the MySQL site.

During the event we will show you why and how MySQL can be used in a data warehouse project. I will cover this part. I will also show a basic cookbook with tips and tricks on how to set up MySQL in a DW environment.

The other speakers will be our partners Infobright and Talend. Infobright will present its columnar storage engine, based on MySQL, which gives incredible performance and flexibility for data warehouses and data marts. Talend will present its ETL solutions and it will show live how the products can be integrated.

This is not the first MySQL Breakfast that we are hosting in London. In November last year we had a HA Breakfast that generated great interest and attracted a large number of attendees. The slides of that event are available here.

I hope to see you at the event!

-ivan

Here We Go Again - MySQL for Online Application Series 2009

2009-08-25T12:11:00.003+01:00

Tomorrow @10 am Central European time we will host the first part of a new series of MySQL for Online Applications - finally!

Registrations and details are here: http://www.mysql.com/news-and-events/web-seminars/online-apps.html

If you are a MySQL and webinar aficionado, you may remember the successful series that we ran in Spring 2007. Two and more years ago it was a great event, with huge participation - it is still available here.

Now we have renewed the material and we have added more hints and information re HA, performance and tuning, online architectures. There is, as usual, some content for developers, for IT managers, for DBAs and for architects. I hope nobody will be disappointed.

This time we will run four webinars:
Part 1 - Introduction to Scale Out and Replication Services (tomorrow, Wed 26th of Aug)
Part 2 - Tuning and Optimising MySQL for Maximum Performance (Wed 9th of Sep)
Part 3 - Advanced Tecniques for High Traffic Sites (Thu 17th of Sep)
Part 4 - High Availability Architectures for Online Applications (Tue 29th of Sep)

I hope to see you all online tomorrow!

Meetup Group in Dublin, 2009-06-24

2009-06-25T10:42:00.004+01:00

Well signposted :)

Thanks to echolibre for providing the content of the beers tables (in the pub schema, of course)!

What future for MySQL customers?

2009-06-24T17:49:00.000+01:00

Yes, that's right, I mean *customers*. What do I mean with "MySQL Customer"? I would say "users who work for companies that prefer to buy products (tech support, subscriptions to enterprise binaries, short consulting engagements etc.)" instead of going for the familiar DIY approach.

Two years ago there was one and only MySQL product. There was the pluggable storage engine API, there was InnoDB outside the core development of the product and nothing else. Customers could confidently refer to a single provider, MySQL AB, to buy services and support from the creators of MySQL. Was it good for the customer? I would say yes, because the product was (and still is, of course) open source and there was a strong ecosystem and a wide adoption. But there was not much choice around: apart from relatively few strong but relatively companies, customers were in some way tied only to one provider.

Last year Sun acquired MySQL AB. I would say that very little changed for current customers, but there were more opportunities to see the product adopted in new sectors. Sun brought a strong credibility for MySQL in the Enterprise. MySQL was then perceived as a serious alternative to more expensive products. I say "then perceived" because MySQL has always been a serious alternative, but not many people outside the standard MySQL comfortable zone were aware or agreed on that. I consider myself lucky because I can visit tens of users every month and I am exposed to all sorts of environments. In this scenario, I soon figured out that the open source world, the web, the social networks, the online solutions is one world; the proprietary world, the enterprise, the legacy systems and the IT departments are another world. These worlds barely communicate, they do not know each other and they rarely share the choices they make. So, back to my point, Sun opened lots of doors (and minds) to the possibility of using MySQL in new projects.

Only two months ago we had the MySQL User Conference. It was a disruptive event - not in a negative or positive way, things simply changed very quickly. The announce of an acquisition at the very same time, lots of internal and external projects (Drizzle, just to name IMO the most important one), more stable and new storage engines, improvements and more important, the Percona conference forced many people to think of the future of MySQL. Few days later, the Open Database Alliance generated even more questions.

The new enterprise customers, the once outside the "comfy zone", seemed confused. They wondered about the future of MySQL, some were optimistic, some pessimistic, very few did not react.

Today the community around MySQL is more active than ever. I think it's fair to say that now we have a real community, in the sense of having lots of contributions, debates and actions. There are at least 20 great projects and products that are worth to mention; these project may bring significant improvements for MySQL users. People deeply involved in the development of MySQL or in contributions are also moving from one company to another and they are bringing knowledge into new and sometimes bigger teams. In open source, it's a non sense to think that a company can loose knowledge. A company may loose the skills or a talented person, but the knowledge will stay forever and the whole community will benefit from it.

So, now MySQL customers have more choice. They can use MySQL as is, they can download it from the MySQL web site, install it and use it. Or they can add/integrate/implement modules and products like DRBD, Infobright, xtradb, memcached, innodb 1.1, PBXT, Waffle, or lots of other patches and features available. They can also look at 5.4 or Drizzle, or MariaDB. And more important, customers can now see two significant changes. First, they have more choice to find the best services - consulting, training, support, subscription services, cloud systems, hosting - on the market, still with a cost that is far way lower than the one around commercial products. Second, with such an active community, they have acquired a free life insurance for MySQL and for the technology around it. There's more than any commercial product can bring on the table of a customer, at any time. And again, there is the great value of open source.

Follow the MySQL Meetup on Twitter tonight!

2009-05-13T14:44:00.004+01:00

If you can't join us tonight for the Meetup event in London, don't
worry, I am sure you will find lots of info and comments on Twitter.

The official tag is #mysqlldn

New Meetup Event in London

2009-05-11T06:14:00.000+01:00

If you live near London or in the South East, mark your calendars! A new meetup event is coming this week.

We will meet @6pm this Wednesday at:

The Lamb
94 Lambs Conduit St
Bloombury
WC1N 3LZ London

From MySQL, there will be Andrew Morgan, Mark Baker and myself.

As usual, there isn't a strict agenda, the main topic will probably be the outcome of the recent conferences in Santa Clara, the MySQL User Conference and the Percona Conference. We have lots of tech stuff to cover - Mysql 5.4, MySQL Cluster 7, Drizzle and Xtradb, just to mention few.

I hope to see you there!

-ivan

P.S. As you know, we cannot comment on the recent Oracle acquisition, so you have these options:
A) don't ask and enjoy the evening
B) buy us few pints and expect non-sense from us :)
C) all of the above

A quick test with MySQL 5.4

2009-05-09T08:04:00.012+01:00

Louis I
Louis II
Louis III
Louis IV
Louis V
Louis VI
Louis VII
Louis VIII
Louis IX
Louis X (dit le Hutin)	(the Quarrelsome)
Louis XI
Louis XII
Louis XIII
Louis XIV
Louis XV
Louis XVI
Louis XVII
Louis XVIII
et plus personne plus rien..	and nobody, nothing more…
qu'est-ce que c'est que ces gens-là	what’s with those people
qui ne sont pas foutus	who can’t even
de compter jusqu'à vingt?	count up to twenty?

Jacques Prevert

I did a quick test with the last version on the same testbed environment that I used for my previous blog and here is my finding.
MySQL 5.4 worked as I expected with sysbench. Performance up to 32 threads are basically the same of my build, but whilst 64 concurrent threads show a degradation in performance of the standard build and my build, with 5.4 they increase significantly, showing a pretty linear scalability - see table and graph.

Threads	5.1.33-Std	5.1.33-IZ	5.4
1	28.63	31.28	28.63
2	58.15	60.21	58.57
4	93.56	108.14	96.57
8	170.46	179.59	166.93
16	286.28	307.78	271.09
32	352.90	400.14	392.68
64	307.62	366.04	576.55
128	222.49	266.61	247.96

I would not recommend to go for 5.4 yet, but the results are extremely promising.

Forgive me, I am back... with 20% more performance on Solaris 10...

2009-04-09T07:26:00.012+01:00

Yes, it's been a long time. Week after week, I had ideas I wanted to share, but no time at all to blog. Priorities are clear: family and work come first, and it's very difficult to keep a balance between the two.

So I left a lot behind. The major project I worked on last summer, MySQL in a virtualised environment and with multiple instances on larger boxes, was part of the material we presented at the MySQL Customer Conference back in October. The results were pretty good and we showed a good level of scalability for real life applications, well beyond the typical sysbench results you can find by googling a bit here and there. I will blog about it at some point, it's just that it takes a lot of time to reorganise all the material to publish and make it understandable.

This week the topic is slightly different. We did it again, we dug into the art of building MySQL binaries and we did it on Solaris 10 for SPARC servers.

The build

Before going any further, a word of caution. My recommendation is always to rely on the certified binaries that our Build team provide, whether they are Community or Enterprise. But if you want to experiment and if you have time to work on extensive testing, then I suggest to have a look at some of the hints you can find on the net. Again, I have to thank my colleagues, Ritu Kamboj and Luojia Chen, for providing invaluable help and support on this matter.

Back to the real stuff. I built a new version of MySQL Enterprise 5.1.32 for a customer by tweaking some config params a bit and we ended up with a boost in performance between 5% and 20% on sysbench and between 18% and 33% on the application testing.

I could use a T2000 SPARC server to build MySQL, running Solaris 10. I used the latest version of SunStudio 12 and I simply changed this:

In the source code, the file .../storage/innobase/include/univ.i should have line 61 changed as:

/* MY CHANGE */
/* # if !defined(__GNUC__) */
# if !defined(__GNUC__) && !defined(__SUNPRO_C)
# undef UNIV_MUST_NOT_INLINE /* Remove compiler warning */
# define UNIV_MUST_NOT_INLINE
# endif

This is the configure command I used:

./configure
'--prefix=/usr/local/mysql'
'--localstatedir=/usr/local/mysql/data'
'--libexecdir=/usr/local/mysql/bin'
'--with-comment=MySQL Enterprise Server - Advanced Edition (GPL) Built by IZ'
'--with-server-suffix=-enterprise-gpl-advanced'
'--enable-thread-safe-client'
'--enable-local-infile'
'--with-mysqld-libs=-lmtmalloc'
'--with-pic'
'--with-client-ldflags=-static'
'--with-mysqld-ldflags=-static'
'--with-zlib-dir=bundled'
'--with-big-tables'
'--with-ssl'
'--with-readline'
'--with-embedded-server'
'--with-plugins=partition,innobase,archive,blackhole,csv,federated'
'--with-extra-charsets=complex'
'CC=/opt/SUNWspro/bin/cc'
'CXX=/opt/SUNWspro/bin/CC'
'CFLAGS=-Xa -fast -xbuiltin=%all -xO4 -xstrconst -mt -m64 -xunroll=2 -xprefetch=auto -xprefetch_level=3'
'LDFLAGS=-m64'
'CPPFLAGS=-D_FORTEC_'
'CXXFLAGS=-fast -xbuiltin=%all -mt -m64 -xO4 -xunroll=2 -xprefetch=auto -xprefetch_level=3 -noex -fsimple=1 -fns=no'

Here are some comments:

CC and CXX refer to your SunStudio 12 compiler, so you should set the right path
Partitioning, InnoDB, Archive, Blackhole, CSV and Federated are available (see '--with-plugins')
Flags -xO4 (optimisation level) and -Xa (only for the C compiler) seem to give a good improvement.

The results

We ran two different sets of tests, one generic and one related to the application that the customer is going to implement.

The first set was based on the popular sysbench. We used InnoDB (since InnoDB is the storage engine that the customer was going to use), measuring read and write performance. The system scaled pretty well up to 64 threads - see table below - reaching 366tps with the new build against 307tps with the original build.


Threads   IZ  Std IZ vs Std
1  31.28  28.63   9.26%
2  60.21  58.15   3.54%
4 108.14  93.56  15.58%
8 179.59 170.46   5.36%
16 307.78 286.28   7.51%
32 400.14 352.90  13.39%
64 366.04 307.62  18.99%
128 266.61 222.49  19.83%

The final output on sysbench with 64 threads on the new build is here:


sysbench
 --test=oltp
 --oltp-table-size=1000000
 --mysql-socket=/tmp/mysql.sock
 --mysql-user=root
 prepare

sysbench
 --num-threads=64
 --max-requests=100000
 --test=oltp
 --oltp-table-size=1000000
 --mysql-socket=/tmp/mysql.sock
 --mysql-user=root
 run

Number of threads: 64

OLTP test statistics:
   queries performed:
       read:                            1402170
       write:                           500775
       other:                           200310
       total:                           2103255
   transactions:                        100155 (366.04 per sec.)
   deadlocks:                           0      (0.00 per sec.)
   read/write requests:                 1902945 (6954.84 per sec.)
   other operations:                    200310 (732.09 per sec.)

Test execution summary:
   total time:                          273.6145s
   total number of events:              100155
   total time taken by event execution: 17481.6191
   per-request statistics:
        min:                                 40.55ms
        avg:                                174.55ms
        max:                               1029.61ms
        approx.  95 percentile:             317.37ms
Threads fairness:
   events (avg/stddev):           1564.9219/21.04
   execution time (avg/stddev):   273.1503/0.04

In the second test, we used a large InnoDB table, with 200M rows, using approx 80GB of space. We used a relatively large buffer size (4 GB), but we measured the performance with a cold cache. This test first creates a large MyISAM temp table from the main InnoDB table, then it appends more rows on the temp table and it scans it to select and to aggregate data. The initial query is based on key access, but the cardinality is not very high and the filter on the SELECT query gets a result of several million rows.
The table below shows the results of the tests. In this test, the queries were executed in a single thread, in a typical transformation for OLTP data on reporting. As you can see, the large scan on the InnoDB tables has a significant improvement of 27% in the new build, whilst the initial sequential scan on the MyISAM table showed an amazing improvement of 33%. The numbers showed in the table are execution times in seconds.


      IZ Build Std Build      IZ vs Std
INS MyISAM TEMP fm InnoDB   4,584.61  5,830.58 27.18%
SEL MAX(Ln)) FROM MyISAM TEMP     898.31  1,194.78 33.00%
2nd INS MyISAM TEMP fm InnoDB   1,343.36  1,590.90 18.43%
SEL Cols and SUM() fm MyISAM TEMP  1,499.29  1,807.32 20.55%
SEL Cols and SUM() fm MyISAM TEMP    884.03  1,105.91 25.10%
SEL Cols and SUM() fm MyISAM TEMP    871.29  1,092.53 25.39%
SEL Cols and SUM() fm MyISAM TEMP  1,516.87  1,817.73 19.83%
SEL Cols and SUM() fm MyISAM TEMP    883.57  1,109.12 25.53%
SEL Cols and SUM() fm MyISAM TEMP    871.01  1,094.52 25.66%
SEL w/WHERE fm MyISAM TEMP   1,406.66  1,637.22 16.39%
SEL w/WHERE fm MyISAM TEMP     890.06  1,114.06 25.17%

   TOTAL (sec) 15,649.06 19,394.67 23.94%

For your reference, here is the MySQL config file, for both tests:

[mysqld]
basedir=/usr/local/mysql
datadir=/usr/local/mysql/data
socket=/tmp/mysql.sock
innodb_locks_unsafe_for_binlog=1
back_log=50
innodb_additional_mem_pool_size=10M
innodb_buffer_pool_size=4G
innodb_flush_method=O_DIRECT
innodb_file_per_table
innodb_file_io_threads=8
innodb_log_buffer_size=6M
innodb_log_file_size=120M
join_buffer_size=8M
key_buffer_size=10M
log_long_format
long_query_time=60
lower_case_table_names=1
max_heap_table_size=16M
query_cache_limit=2M
query_cache_size=64M
read_buffer_size=4M
read_rnd_buffer_size=10M
sort_buffer_size=2M
table_cache=32
thread_cache_size=8
tmp_table_size=16M

101 MySQL Stories - 5.1: Is that So?

2008-12-09T22:01:00.004Z

Rick Deckard is an Engineering Manager at MySQL. Rick is one of the most senior engineers in the company, he led the team during the long delivery of the newest GA version of the server, 5.1.

Soon after the release of 5.1, the Internet Village experienced shocking hours. A wind of disappointment spread all over the Net, blown by news of serious issues that affected 5.1. Many users were concerned, they feared for their mission critical applications based on 5.1, although they had already been in production for months without a glitch.

With great disappointment, the Net pointed the finger at Rick, asking for a clarification on these serious issues. "Is that so?" was all he would say.

After a few days, some wise users spent time reviewing the issues that affected 5.1. They discovered that some issues were not different by the ones that were affecting 5.0, and that version was currently in production. The ones who had already deployed 5.1, did not experience any particular problem. On the contrary, the servers were dealing pretty well with the increasing number of queries of some financial sites.

Many users also did some math on bug counting. A user discovered that the 29 releases of the most robust and simple version 4.1, was affected by 887 S1 and S2 bugs and 2 of them are still unresolved. For 5.1, they found only 193 S1 and S2 bugs in the first 30 releases, although 10 are still unresolved.

Wise men in large enterprises brought up the subject in their technical committees. Unanimously, DB Architects and DB Administrators agreed that 10 S1/S2 bugs in the first GA version of a commercial database is a dream that will never come true. Several organisations decided to expand the use MySQL as a result of this due diligence.

Few weeks later, Rick Deckard was invited to a Meetup group in the Valley; the main topic was "The great new features of 5.1 and the roadmap for the next releases". The meetup was a great success. After the meeting, some users approached Rick and showed him some graphs produced with the new Enterprise Monitor 2.0 and the Query Analyzer, showing a significant improvement in performance and scalability with some new features of 5.1: partitioning, row based replication, extended use of memory for caching.

With the graphs in his hands, all Rick said was: "Is that so?"

This story is fictitous. Any reference to persons and/or MySQL Customers is purely coincidental. But this is not far off from one day in the life of a MySQL Sales Engineer.

Join me and Ralf Gebhardt in the web seminar "MySQL 5.1", Thursday 11th of December at 10am GMT - http://www.mysql.com/news-and-events/web-seminars/display-251.html

Let's find out all you need to know about 5.1!

Final Countdown - MySQL European Conferences

2008-10-16T00:18:00.001+01:00

A MySQL Conference in Europe? Wrong: three MySQL Conferences in Europe!

Yes, this is the plan for next week - 21st of October in Munich and 23rd in London - and for mid November - 19th of November in Paris.

Each venue will host a full day of presentations, news and analysis on MySQL. You can have a look at the official web page for all the details, agenda and registration.

The content can satisfy the appetite of a technical audience and of more business-oriented IT managers at the same time. We will have sessions on performance, on scalability solutions and on Proxy, with hands-on the servers, difference parameters and tools. IT managers will probably find interesting a renewed set of HA solutions and some renewed views on the infrastructures used to power the Web.

Robin Schumacher, Director of Product Management, will present the new roadmap for our products and Miriam Tuerk, CEO of Infobright, will talk about this impressive open source, column-based storage engine, specifically designed for data warehouse and business intelligence applications.

As a final keynote, we will present a set of benchmarks of how MySQL Server behaves and scales in different environments and with different scaling strategies. It's new material, something that we have never presented before, all unveiled next week.

Stay tuned!

Tip of the Day - Queries that you do not want to replicate

2008-10-15T23:41:00.002+01:00

If you use MySQL Replication, 99% of the times your main concern is to keep the master and the slave fully consistent and in sync. Some applications require exactly the opposite.

Let's take, for example, some reporting solutions. You may want to run reports against a data mart (ok, I can see some BI experts here jumping off their chairs - please forgive the simplification for the sake of this example), that is a subset of the whole historical data within your data warehouse. So, you may have a master server containing only the last 3 months of your sales and a slave containing three years of sales. Sales managers can happily to see current year/previous year rolling comparisons using the slave server and at the same time they can run faster current month/previous month reports on their master.

There are many ways to prepare the data mart and the historical DB. The easiest thing to do is to remove the oldest rows from the data mart, but not from the historical DB. If you use partitions (out of the box with 5.1 or with application partitioning in 5.0), for example based on a date column, you can drop a full partition and speed up the whole process. So, it does not matter how you remove your oldest rows from the data mart (here, the master). What it matters is that you do not want to replicate these commands to your historical DB (here, the slave).

Here is a simple trick. Let's assume that the schema that contains the data is called COMPANY_SALES. All you have to do is to create a new schema called REPL_ADMIN (or a name that you may fancy more) and add the binlog-ignore-db parameter in your server config file. The my.cnf on your master will look like:

[mysqld]
...
binlog-ignore-db = REPL_ADMIN
...

Ok, now the real trick. Every time you want to remove the rows from a sales table, simply "move" to the admin schema using the USE command first. Something like this works pretty well:

USE REPL_ADMIN
DELETE FROM COMPANY_SALES.fact_sales
WHERE sales_date <= DATE_SUB( CURRENT_DATE, INTERVAL 3 MONTH )

With the DELETE command, you remove all the rows order than 2 months from the table fact_sales in the schema COMPANY_SALES. Since you used USE and you "moved" to the REPL_ADMIN schema first, the binary log will not keep track of the DELETE, even if it affects the COMPANY_SALES schema.

Some people refer to this behaviour as a bug. Personally I find it extremely powerful: it gives replication an incredible flexibility and the ability to do what other replication technologies can't do.

Tip of the Day: Installing MySQL Coolstack on Solaris

2008-10-11T11:36:00.002+01:00

Here are few tips if you are going to install the MySQL Coolstack package on Sun Solaris.

First of all, let me say that, as usual, these tips are based on my personal experience and they may not be applicable to your sw version or environment. For this specific tip, I can't say if it is applicable for all versions of Solaris - 9, 10, OpenSolaris, X86, SPARC etc.

You can get the Coolstack from here - http://cooltools.sunsource.net/coolstack. Within Coolstack you will find CSKmysql, a specific 64 bit package prepared by Sun that provides some interesting performance improvements. Coolstack has been built with Sun Studio from the same source used to build the standard versions. Also, CSKmysql provides Dtrace probes - I am sure some will find this feature a great plus. The last available version, 1.3.1, is based on MySQL 5.1.25RC.

When you install the package, make sure that you first create a symbolic link to your package in /var/sfw/lib/64, in order to avoid the message:

pkgadd: ERROR: no packages were found in

The package is installed in /opt/coolstack/mysql. Unless it does impact your internal policies or other software, I would recommend to create the symbolic link /usr/local/mysql.

Once the software has been installed, you need to create the first set of system tables. For some reason, the mysq_install_db script (which is usually in the scripts dir) is not available in Coolstack. You need to download a community edition from the mysql web site or you may create the tables on another server (Linux or Solaris, it does not matter) and then copy them on the server you are preparing. If you decide for the latter, make sure you apply the right owner (usually mysql).

You are almost there. The last issue you may find is some missing libraries. The easiest thing to do is to export the LD_LIBRARY_PATH var:

LD_LIBRARY_PATH=/usr/sfw/lib/64

Make sure you set the LD_LIBRARY_PATH for both your server and your client.

Et voila'. The server is (should be) able to start with the usual mysqld_safe command.

MySQL at the FOWA - Future Of Web Applications - in London, 8-10 Oct

2008-10-04T21:37:00.003+01:00

I know, it's a very short notice. But if you are in or around London next week, you might want to come and visit us at the FOWA Expo. The venue is the London ExCel; some of you should know it, since it's a quite popular that hosts IT and social events.

Check the FOWA Website - http://london2008.futureofwebapps.com - for more details. The schedule is pretty packed with speakers from Amazon, Google, Digg, AOL, SalesForce, BT and many others... and of course from Sun.

As MySQL, we do not have any speech scheduled, but we have a booth and if you have any question or you want to get in touch with us, you are more than welcome to stop by! This time we promise the booth will be less minimalist - check what we had last year :) - but you know us: no frills and lots of hard work at a back-end...

I hope to see you there!

/iz

101 MySQL Stories

2008-08-09T20:07:00.000+01:00

Hi,

My name is Larry and I am an Oracle DBA. I work in the City for a large finance corporation - we use Oracle, SQLServer and Sybase. It's a great job, very well paid and plenty of opportunities to learn new stuff.

My company has been seriously hit by the crisis in the financial sector and the management is now wondering if it's really necessary to pay expensive licenses for every small project that requires a database. It's true that we have a corporate license agreement, but that does not cover all the instances - we need to distinguish between front end, back end, sockets, physical CPUs, cores, threads and lots of other rules I do not remember.

One of my managers heard about MySQL. He read that MySQL powers the majority of the Web 2.0 world. I thought that MySQL was only a small thing for home use, but maybe it's been improved. I am curious by nature, so I am really excited to discover something new. I am not sure whether a MySQL DBA is paid as much as an Oracle DBA - I'd better Monster a little bit to check - but I am not that greedy and I am sure there are opportunities for MySQL DBAs as well.

Today I am going to meet Inn-oh, apparently a well known MySQL Sales Engineer who moved from Tokyo to London just recently. We will meet at the Solution Centre in the City - he is certainly going to show me massive servers with large instances of a new version of MySQL.

I am really excited. I want to know how MySQL can handle such large web sites. I am sure they will use something like Oracle RAC, maybe a large Storage Area Network from Sun, all active servers, all in a single huge rack with lots of spindles. Still, I do not know how they can manage the thousands of connect/disconnect to the DB every second - programmers really make our life difficult!

I want to know everything about the MySQL Stored Procedure Language. Our company applications heavily rely on PL/SQL; we have hundreds of thousands lines of code. Oracle strongly recommended it for performance reasons, but now I don't know how easy it will be to migrate to another database. Within the database instance we do everything: we send emails, we build web pages, we have even built a full CMS, all in PL/SQL!

I am now at the Sun Solution Centre, Inn-oh is waiting for me. He asks me if I want a drink, I'd love a cup of tea. We move to the table where there are a couple of kettles and some cups. Inn-oh pours the tea in the cup. The cup is full, but inn-oh keeps pouring. "I am sorry Inn-oh" I say, "but the cup is already full, "you can't put more tea in it!".

"Like this cup,", Inn-oh replied, "you are full of your own opinions on how to perform scalability and high availability. You've been told that the database server must be the centre of every system, that complex data and business logic must reside in it. You believe that a single, non-splittable element (the SAN) can provide unlimited scalability and reliability. How can I show you how MySQL scales and performs unless you first empty your cup?".

This story is fictitous. Any reference to persons and/or Oracle or MySQL Customers is purely coincidental. But this is not far off from one day in the life of a MySQL Sales Engineer.

Drizzle - Here Comes The Rain Again (which might be good for Enterprise users)

2008-08-03T08:42:00.001+01:00

When I moved to the UK from Italy, now more than eight years ago, people asked me all the time "Why did you leave such a sunny place like Italy to move to England?". Well, first of all, the south east of England is not that bad. There is a decent number of sunny days in a year and the bright sky of a breezy morning is just fantastic. Second, my home town is Milan, in the middle of the Po valley, with foggy winters and humid summers, with temperatures between -15C and +38C over the year. And there is another point: no wind and no much rain, hence pollution. In Milan the pollution is so awful that almost one third of the working days in a year private cars cannot be used within the city, in the attempt of reducing emissions. When I was a child, my mum used to put a PVC cover over the washing line to avoid that clean clothes would get immediately dirty.

It is a long introduction, but it is probably worth it, because it fits almost perfectly with the new project you may find in the MySQL pot - Drizzle.

Drizzle is a Community project led by some of my colleagues - Brian Aker, Jay Pipes, Monty Taylor among many others - aiming at building a microkernel-like version of MySQL, without jeopardising the intrinsic nature of the software. In fact, Drizzle is based on the robust code around MySQL 4.1, the last version developed before MySQL decided to move to the Enterprise market and added lots of new Enterprise-like features like stored procedures, triggers, views etc.

First of all, nothing is going to change in the current version of MySQL. 5.0, 5.1 and 6.0 are still there and there is no plan at the moment to apply any change to the core or to any of the modules within the server. Drizzle is really a new project, far from being 100% ready for production for everybody (i.e. unless you are ready to jump in and look at possible issues within the code). The community has welcomed the project enthusiastically and, after few days and a warning alert that I keep in my mind at low priority, I am going to join the enthusiasts.

The first objective of Drizzle is to provide the essential features of a database server used in a typical web environment. Drizzle has been designed (well, it is still in a design phase) with Web 2.0 in mind. The DB must be fast, reliable and not many features are really needed. This is just fantastic, because it may open many possibilities. But Drizzle will be even more interesting when the Enterprise as a whole - meaning any kind of applications with a Web front end - may take advantage from it. The greatest value of Drizzle would be in fact a reorganisation and modularisation of the code. It may be, in some way, like my move from a sunny-but-polluted part of Italy (where stopping traffic helped only a bit), to a breezy land with less sunny days, but with fresh air and stunning landscapes.

Once (if) Drizzle will be at some point stable in its microkernel structure, then the Community and Sun will start adding features through plugins, not messing up the kernel - whenever possible. You may think about some features available now in 5.X that are not available in Drizzle. Triggers, views and stored procedures, just to name the three most common enterprise features, may all be added outside the kernel and they can be switched on when needed. Think about the infinite possibilities here. With stored procedures, for example, you may have different plugins to choose from the SP language you want to use. Triggers can be used in a object-relational fashion to capture reads and not only writes, when needed. Foreign keys and referential integrity will finally be outside the storage engine as they should be.

Is Drizzle good for the Community? Definitely! Because it gives anybody more options to integrate new features that are not available at the moment and that are not part of an immediate roadmap at Sun/MySQL.

Is Drizzle good for the Enterprise? It can be! Some people think it's too difficult to build a modular server with the current core, some others think this is not easy, but possible and preferable. There is an open debate and this is a great start.

So far, so good. So, what's the warning? Well, we at Sun have a duty with our users and customers - we must deliver reliable, high performance and ease of use versions of our products (not just the server). I really hope that Drizzle will not move the focus of our key contributors, slowing down bug fixing and performance improvements. If Drizzle can remain a parallel project and at some point will be good enough to become the mainstream product (something like the change from Mac OS to Mac OS X for Apple) with all the required plugins for the millions of installations out there, then we will finally provide the best database of the third millennium.

Go Drizzle!

Italian Webinar - Materiale, Domande e Risposte per il Webinar "Guida alla scalabilita' di MySQL"

2008-06-13T18:18:00.003+01:00

La registrazione del webinar e le slide sono presenti qui.

Di seguito riporto la sezione di domande e risposte:

Q from Ivo: Tutte le considerazioni sulla scalabilita' fin qui individuate sono valide anche per la versione community di MySQL?
A: si

Q from Alex: Mysql Proxy per la suddivisione delle richieste lettura scittura è ad oggi stabile ?
A: MySQL Proxy e' ancora in fase alfa e non e' consigliato l'utilizzo in produzione.

Q from Massimo: Qual'e' il metodo migliore per il backup incrementale di un db?
A: E' necessario attivara il binary log ed effettuare le operazioni di FLUSH LOG al momento del backup.

Q from Mauro: Ci possono essere applicativi web che inseriscono dei dati nel DB, e che subito rileggono i dati inseriti. Dato che la replicazione è asincrona, non si ha la sicurezza che uno slave abbia subito l'aggiornamento. Quale soluzioni si possono trovare per questo?
A: La soluzione migliore e' quella di operare le letture che necessitano di una sicurezza del dato letto direttamente dal master. Questa necessita' si ha anche nel momento in cui le letture sono isolate all'interno di una specifica transazione.

Q from Fabio: Con un traffico di circa 4 milioni di pagine al mese un server slave ben configurato può bastare?
A: E' difficile fare valutazioni su queste basi. Dipende molto dalle query effettuate per produrre queste pagine, nonche' le richieste di picco, piu' che un valore totale per mese.

Q from Gianni: Qual'è l'ordine di grandezza del ritardo nella replicazione asincrona da singolo Master a singolo Slave: secondi, minuti, ore?
A: La replicazione puo' essere (o e' auspicabile che sia) istantanea, il tutto dipende dal carico di lavoro dei sistemi. Nel caso di sistemi molto carichi, il throughput richiesto potrebbe essere superiore a quello supportato, nel qual caso si registrano i ritardi citati.

Q from Dan: Per la replicazione in lettura dovrei installare piu' servers mysql? Dove posso trovare degli approfondimenti?
A: Si, la replicazione si basa su piu' server MySQL connessi in rete tra loro. Il manuale riferimento pubblicato sul sito fornisce tutti i dettagli del caso.

Q from Francesco: che database usa Ebay?
A: Ebay utilizza MySQL per gestire le sessioni online. Nello specifico, si basa sull'utilizzo del MEMORY storage engine.

Q from Gianluca: Quale è lultima versione di MySQL ent certificata per Linux Red HAT Advanced Server 5?
A: Alla data della mia risposta (10 Giugno 2008) e' la 5.0.62

Q from Stefano: non si otterrebbe maggiore scalabilità inserendo i dati a bassa persistenza sul master?
A: Il problema e' quello di garantire che le risorse a disposizione non siano un collo di bottiglia. Se utilizziamo il master per memorizzare dati a bassa persistenza, questo normalmente significa che effettueremo una notevole quantita' di update, quindi genereremo una notevole quantita' di scritture su una risorsa che e' normalmente piu' difficile da scalare. Inoltre, se i dati sono a bassa persistenza, ha poco senso che venganon costantemente replicati, in quanto ci si aspetta che la loro lettura debba essere svolta senza ritardi e la validita' del dato sia molto breve.

Q from Davide: Installando il Community Server non ho visto nessuna sezione per creare gli slave dipendenti. C'e' un opzione che non ho visto o devo prendere degli addon?
A: No, la replicazione e' disponibile nel Community Server. Per impostare la replicazione e' pero' necessario eseguire alcuni comandi di amministrazione, non e' una cosa che avviene a livello di installazione.

Q from Fabio: Nel caso sharding, se separiamo logicamente le tabelle o addirittura gli schema, e' possibile usare le tabelle federated per fare join? Consigliato performante o e' meglio replicare tra i master le eventuali tabelle utili alle join?
A: E' possibile utilizzare delle tabelle FEDERATED, ma generalmente le prestazioni ottenute non sono entusiasmanti. Consiglierei piu' una replicazione, se possibile.

Q from Meo: MySQL Proxy mi risulta ancora in versione Alpha. Vi sono previsioni per un suo prossimo rilascio come GA? Vi sara' anche una edizione enterprise?
A: Non sappiamo ancora quando sara' GA. E' importante considerare che i prodotti in alfa possono essere soggetti a sostanziali modifiche. E' probabile che Proxy sara' supportato con MySQL Enterprise.

Q from Marco: Per utilizzare mysql proxy bisogna scrivere del codice SQL con istruzioni specifiche? ( sul nodo MASTER o SLAVE DOVE effettuare l'operazione ) ?
A: Questo dipende da come Proxy acquisisce e gestisce la connessione proveniente dal client. La valutazione puo' essere svolta all'interno del Proxy tramite degli script generici preimpostati, come avviene in alcuni esempi che accompagnano proxy.

Q from Aldo: Avete un sistema backup per il cluster mysql?
A: Assolutamente. All'interno di MySQL Cluster e' presente un sistema di backup e restore online.

Q from Marcello: Per una applicazione web con moltissime letture e poche scritture, che stiamo progettando con Cluster di tutte le tabelle e Replicazione delle sole tabelle di lettura, è buona norma che le tabelle del cluster siano innoDB e quelle replicate MyISAM?
A: Intendo quindi che per "cluster" si intende cluster a livello di sistema operativo. Il cambio di storage engine e' possibile. Va valutato se per l'applicazione specifica porta eventuali benefici. Un'altra possibilita' da valutare e' di velocizzare innodb, ad esempio riducendo il numero di sync su disco oppure adottando una diversa strategia RAID.

Q from Aldo: Per garantire l'alta affidabilita' di mysql cluster quali sono i processi che mi devo prendere carico io come applicativo di restartare se vanno giu? esempio managemnet node e sqlnode e basta o solo management node?
A: Se uno dei componenti ha subito un fault, non e' necessario farlo ripartire, e' opportuno investigare qualie problema ha incontrato. Sconsiglierei in ogni caso la ripartenza da applicativo.

Q from Alessandro: Che grado di stabilità hanno la versione 5.1 e la 6.0 di mysql?
A: La versione 5.1 e' prossima ad essere rilasciata GA, quindi la sua stabilita' e' quasi da GA. La versione 6.0 e' ancora in fase alfa, quindi sconsiglio vivamente il suo utilizzo in produzione.

Q from Fabio: Esistono benchmark che dato un determinato tipo di hardware, mostrano le prestazioni di varie tipologie di soluzioni di scalabilità? es query al secondo ecc.
A: Ci sono alcuni benchmark presenti sul nostro sito (basta cercare la voce "benchmark"). I siti dei fornitori di HW sono sicuramente piu' forniti sull'argomento.

Q from Alessandro: Per quanto riguarda gli esami di certificazione di livello dba (non mca) sono obbligatori i corsi in aula?
A: No, non c'e' nessun obbligo di partecipazione.

Q from Aldo: Il backup incrementale c'e' anche per il cluster mysql?
A: No, MySQL Cluster supporta solo il backup a livello di intero nodo.

Q from Mauro: Quali sistemi di heartbeat per la replication esistono?
A: Si puo' usare un qualsiasi sistema di clustering, come ad esempio Linux Hearbeat, RH Cluster o altri.

Q from Alex: La replicazione circolare Master/slave con Mysql è attuabile in un ambiente di produzione? Se si con quanti nodi max?
A: E' attuabile e non vi e' un numero massimo di nodi. E' pero' necessario valutare quale e' la necessita' che porta all'utilizzo della replicaione circolare. L'implementazione di questa topologia non e' infatti scevra da alcuni aspetti da considerare attentamente.

Q from Aldo: ma è possibile impostare una ATTESA per la scrittura? Mi spiego meglio posso impostare il ritorno al mio software SOLO DOPO che la write è andata su tutti gli slave?
A: No, auesto non e' possibile. Lo slave viene aggiornato in modo asincrono.

Q from Ivo: Ma la replicazione e gestita completamente da mysql oppure c'e bisogno di qualche applicativo ad hoc da implementare?
A: Non occorre nessun applicativo, e' sufficiente avere almeno due server MySQL

Q from Andrea: Se utilizzo un hardware SAN, posso fare a meno di drbd se voglio realizzare un cluster attivo /passivo?
A: DRBD e' un alternativa a un cluster basato su SAN. E' una soluzione nettamente piu' economica che non richiede l'acquisto di hardware specializzato.

Q from Silvia: Il driver JDBC mysql configurato in "replication mode" che funzionalità di distribuzione del carico fornisce?
A: Permette di suddividere le letture tra i vari slave in modalita' round robin, random o tramite un QoS. Statisticamente, la modalita' random e' la piu' efficiente.

Q&A and Recording of the Memcached Webinar

2008-06-08T12:42:00.006+01:00

The session has been recorded. The webex file and the webinar are available here.

Here is the Q&A session, with more detailed answers.

Q from Tomasz: You said memcached is integrated with, among others, MySQL. Did you mean integration by e.g. php or direct, automatic integration?
A: Memcached API are available for many enviroments, such as Php, Perl, C/C++, Rails etc. With MySQL, you can use UDF functions or a storage engine. Both are community projects.

Q from Juan: memcached: Facts data size up to 1MB? What size, i dont understand!
A: It's the size of an item associated to a key. When you store an element (item) into memcached, the size of the key is max 250 bytes and the item is max 1MByte.

Q from Frederik: Do memcached save data in RAM or disk?
A: It is RAM based, there is no durability

Q from Niko: Are all these [client] libraries available for Windows as well ?
A: Some of those libraries have been ported to Windows

Q from Robin: Is it possible to define multiple memcached nodes? If so, what is the hashing strategy?
A: Yes, it is. Memcached is distributed by design. The distribution is based on the hash key, depending on the number of servers in the cluster. So, for example, in a cluster with two servers, the item with key 1 will go to server A, key 2 will go to server B, key 3 to server A, key 4 to server B and so forth.

Q from Aron: If i have set data into memcache from db, but after a while the db table changes. How do i trigger to update the memcache? MySQL cache deletes data from cache when a table has changed, how does memcache handle this?
A: There are various mechanisms. First of all, memcached is a generic caching mechanism, therefore you can use it to cache data from the DB, but also HTML pages, images, documents etc. There are various mechanisms to trigger an update. One is to apply a trigger to the table in the database and invalidate/update the cache at any change. Other options may be application specific.

Q from Ran: What are the advantages of using memcached via MySQL, as opposed to directly using the memcached API? Can you expand on using triggers to automatically put/get data from cache?
A: Yes, see my previous answer. In general, if your are going to store DB data, it's a good idea to set this kind of trigger.

Q from Jeroen: How can you store the same item on multiple servers?
A: You simply apply two hash keys that will go to different servers. The typical approach is a key on one server and its next value on another server. Be careful to not use the next value key for another item.

Q from Giulio: Can I setup mysql slave to replicate memcached data?
A: This is a very good approach. If you are not specifically worried by the real time update of the cache, you can create triggers on the slave server (and perhaps you can use a BLACKHOLE storage engine to save storage. In this way, the master will not be affected by any delay caused by the triggers.

Q from Matt: Can memcached be installed as a Debian Package without any configuration needed? i.e. "apt-get install memcached"
A: As far as I know, only the client library can be installed as a Debian package.

Q from Ariel: What about network latency if I have so many Memcached servers?
A: The protocol used to interact with memcached is very lightweight. Naturally, the more you use memcached, the more traffic you will generate. So far, none of the very large installations using memcached have had latency problems.

Q from Eyal: How do you define the structure of the items you store?
A: Simply you don't. Memcached will store a string of bytes and it's up to the application to interpret the content of memcached.

Q from Gregorio: Do Facebook or any of them use any kind of integration between Mysql + Memcached?
A: I am not sure about Facebook, but many web sites do. Fotolog, for example, stores DB data using UDF.

Q from Ariel: Is memcached good for queries that take less 1s to respond?
A: the longer the query takes to respond, the more effective is memcached compared to the execution of the query. If you run several queries that take one second or more, maybe it's a good idea to use memcached.

Q from Marc: How do you handle the slab sizes if they are growing too much and using full size of memory? Also when items expire the slab size stays the same. that's a common problem in memecached.
A: Indeed, this is a know problem. Some optimisation in the use of the storage (memory) has been sacrificed in order to keep the algorithm for storage and eviction pretty simple.

Q from Yariv: a question about memcached distribution, how does it work, i understood that you implement distribution in the code using module (%) maybe i'm wrong, but how does it work when nodes are added or removed from the 'memcached ' cluster?
A: If you add or remove nodes, you end up invalidating the whole cache. Bear in mind that we are talking about a caching mechanism, very fast and simple, therefore with basic capabilities in terms of administration.

Q from Varaprasad: Can I use memcached with MS SQL Server?
A: You can, sure, although I am not aware of any connector or function to integrate MS SQL Server to memcached.

Q from Oren: Isn't the cache purpose to offload complexity from the db? Why then to use the mysql memcached functions, and not do it through the application, bypassing the db?
A: It's not just complexity, it's mainly workload. If you handle a query through memcached instead of engaging the storage engine and the consequent I/O, you still have an advantage. I agree with that if you avoid to involve the DB server completely, it's even better.

Q from Varaprasad: Is memcached opensource?
A: Yes, but as you just see on the slide we will support it under the MySQL Enterprise Gold and Platinum subscription.

Q from Pedro: in the MySQL Cluster is there a limit on the database size?
A: With Version 5.0 we are limited to the available RAM of the available data nodes. With 5.1, you can extend the use of the data on disk.

Q from Paolo: Seems like Memcached can be redundant and HA with little effort, is it right?
A: Yes, sort of. You need to implement the redundancy yourself and in any case it's not so simple to guarantee that you will have consistent data on all the servers.

Q from Jan: All memcached nodes can have a different amount of memory reserved for memcached, am I right? Any best practices?
A: That is correct. The servers with more memory will have less chance to evict old data. Recommendation is to use the same amount of memory for memcached server of the same type, i.e. when the server are used to store the same type of data.

Q from Andrew: Is memcached supported as a storage engine or as a standard memcached instance?
A: It is supported as a standard memcached instance. The storage engine is still at an early stage.

Q from Matz: Consistency? How does the cache know when the data has been updated?
A: It doesn't. The DB through a trigger or the application will inform memcached that the cache must be updated.

Q from Ariel: Is there some library for use Memcached in HA Cluster?
A: Not directly, as far as I know.

Q from Juan: How can I backup memcached?
A: There isn't a simple way to do that and usually it's no a good idea. Memcached is designed to provide high performance and you do not want to slow down the server for a backup. Even more important, cache is temporary data by definition, data that you may or may not find in the cache, so there is no point to backup a cache.

Q from Julian: How does memcached scale beyond a single server? The slide said the size was unlimited..
A: You simply need to define a set of servers to use at a client level and the API will distribute the data on these servers.

Q from John: Does memcached work with DRBD?
A: No, they are different technologies. DRBD replicates file system data, while memcached use RAM to store data.

Q from João: Is memcached available in the free mysql version?
A: memcached is available here: http://www.danga.com/memcached . You can use it with our Community Edition, but of course we do not support it with that edition.

Q from Robin: Are there any reports available yet on the performance gains of using memcached from within mysqld?
A: Not directly from MySQL, but there are many tests published on the Web.

Q from Agoston: Can you give an example on how to append timestamp to a memcached key to achieve automatic invalidation?
A: The automatic invalidation is handled only with the LRU algorithm, however you can handle invalidation programmatically. For example, you can defined a version id to append to your key (Something like "mykey-0001", "mykey-0002" and so on). Another option is to prepend a timestamp to the item; you will retrieve the item but you will not use it if the timestamp is outdated.

Q from David: How frequent the data get updated in memcache once DB record has changed? Is it configurable?
A: The update must be explicit, the client or the server through the UDF should update memcached.

Q from Niranjan: Is there some inherent pooling in memcache if I decide to distribute my cache across different machines ? Or will I have to route requests manually ?
A: The requests are directed to the right server using the key, you do not need to route it manually.

Q from Tristan: At what amount of traffic do you think memcache starts being a good option?
A: It really depends on the requirements than on the traffic. Some applications need real time or near real time performance, some others may accept longer response time. This requirements, combined with the concurrent users, should give you an idea whether it makes sense or not to use memcached.

Italian Event: Marten Mickos in Rome, Friday May 30th

2008-05-24T08:38:00.002+01:00

Marten Mickos sara' presente a un evento presso l'Universita' La Sapienza di Roma, Venerdi 30 Maggio.

Marten parlera' della sostenibilita' del modello Open Source, del presente e del futuro di MySQL.

L'agenda e il modulo di iscrizione, per chi fosse interessato, e' disponibile qui.

In coda all'intervento di Marten, parleremo di scalabilita' e compareremo l'approccio alle architetture basate su MySQL rispetto a quelle utilizzate con altri database relazionali.

Giuseppe Maxia illustrera' il ruolo della Community nell'eco-sistema di MySQL e spieghera' come e' possibile contribuire al miglioramento dei prodotti e delle soluzioni.

Ci sara' anche spazio per un'ampia sezione di domande e risposte... meglio approfittarne!

Spero di vedervi numerosi a Roma!

-ivan

New webinar on MySQL and memcached

2008-05-24T08:34:00.001+01:00

Are you ready for the next webinar? This time we will talk about the use of MySQL with memcached. Most of you may already know memcached and how to use it in a typical web environment with high traffic and requirements for high scalability.

During the webinar we will cover different scenarios and best practices regarding the use of MySQL and memcached together.

The webinar will be live next week, Wednesday 28th of May @ 9am GMTD (London, Dublin, Lisbon) / 10am CETD (Paris, Madrid, Munich, Rome etc.)

The registration is available here.

I hope to see you all online next week!
-ivan

Italian Webinar - Materiale, Domande e Risposte per il Webinar "Soluzioni MySQL per l'alta disponibilita'"

2008-05-24T08:25:00.004+01:00

Le slide e la registrazione del webinar che si e' tenuto il 7 Maggio scorso, sono disponibili qui.

Di seguito ho preparato la sessione di domande e risposte.

Q from Andrea: E' possibile creare un sistema multi-master tramite l'assegnazione di impostazioni auto_increment_offset diversi?
A: Si, questo e' il meccanismo piu' tipico. Ogni server master utilizza un diverso offset per creare una PK e quindi inserire nuove righe nella stessa tabella, quindi

Q from Massimo: Cosa succede se viene effettuata una lettura su uno slave di alcuni records modificati nel master e non ancora replicati? si puo prevenire questa situazione?
A: La situazione si puo' prevenire, ma la logica da applicare e' abbastanza complessa. Detto questo, se le specifiche richiedono questo tipo di lettura aggiornata in real time, l'utilizzo della replicazione non e' consigliabile. Le soluzioni che posso ipotizzare sono due:
1. L'utilizzo di due connessioni, una per la lettura su server replicati e l'altra per la scrittura e la lettura di dati che devono essere aggiornati in realtime. Quest'ultima connessione viene indirizzata sul master
2. L'utilizzo di una struttura a storage condiviso in configurazione active-active, qualora l'utilizzo di MyISAM abbia senso (active-active e' possibile solo con MyISAM).

Q from Angelo: Chi decide a chi demandare eventuali query verso i servers slave?
A: La scelta deve essere definita a livello applicativo, utilizzando una logica che e' parte integrante dell'applicazione, uno strato logico di gestione delle connessioni o in alcuni casi un middle layer come ad esempio MySQL Proxy.

Q from Andrea: Il failback automatico può funzionare solo con due server? o per la circolarità della struttura è possibile definire dinamicamente i server che compongono l'anello?
A: La circolarita' puo' essere cambiata, ma l'operazione e' sempre manuale (ovvero non esistono automatismi gia' pronti all'uso).

Q from Angelo: Se io ho un database che occupa 8GB, devo avere 8GB di RAM su ENNE nodi che fanno parte di Mysql Cluster?
A: No, la formula e' la seguente: somma della memoria disponibile in tutti i nodi diviso il numero di repliche (normalmente 2).
Per esempio: 4 nodi con 8GB disponibili su ciascun nodo = 32GB, diviso 2 (2 copie dei dati) = 16GB di memoria disponibile per i dati e gli indici.

Q from Massimo: Le connessioni ai diversi SQL node , come possono essere gestite in maniera trasparente all'utente? Devo per forza passare per Mysql Proxy o esistono altre soluzioni che non siano dipendenti da un applicativo?
A: Non ci sono meccanismi automatici, in quanto un eventuale logica gestita a livello di connessione dovrebbe avere conoscenza del tipo di transazione da eseguire - infatti cio' dovrebbe avvenire tramite Proxy e per tale ragione la soluzione assolutamente piu' sicura e' quella di gestire questa suddivisione a livello applicativo.

Q from Eleonora: Usando DRBD c'è possibilità di perdita di dati nel caso in cui il master cade prima di replicare i dati sul nodo slave?
A: No, questo non puo' avvenire, in quanto la replicazione e' sincrona. Solo cambiando i parametri di configurazione e abbassando il livello di sicurezza della replicazione c'e' il rischio di perdita dei dati - ma ovviamente in una configurazione HA questo e' sconsigliato.

Q from Angelo: Durante il failover, la connessione e quindi la query eventuale, viene persa?
A: Si, le connessioni aperte e le conseguenti transazioni aperte vengono interrotte. Le transazioni aperte non vengono scritte su disco, mantenendo cosi' consistente lo stato del DB.

Q from Massimo: Si può usare il sistema DRDB su macchine virtuali?
A: E' possibile, ma occorre essere certi che a livello hardware non siano presenti eventuali Single Point Of Failure.

Q from Stefano: Nella replica master-master, se cade il master active e successivamente viene riattivato, la risincronizzazione è automatica o richiede operazioni manuali?
A: La sincronizzazione e' automatica. L'unica accortezza e' quella di verificare, prima di attivare il meccanismo di risincronizzazione, che tutte le operazioni siano state trasferite dalla macchina "master active" alla macchina "master passive" prima del fault.

Q from Alessandro: MySql Replication e MySQL Cluster possono essere implementati anche in sistemi windows?
A: La replicazione e' disponibile anche su Windows, MySQL Cluster e' disponibile solo su sistemi Solaris, Linux, OS X, FreeBSD e Windriver.

Q from Alberto: MySQL Proxy non e' considerato un servizio HA?
A: Non nella sua funzionalita' di base e non ancora, essendo ancora in una fase alfa. Tramite l'implementazione di appositi script, verra' sicuramente utilizzato anche per operazioni di failover automatico.

Q from Marco: Posso fare la replica con due istanze che risiedono su server con sistemi operativi differenti? Ad esempio Windows e Linux
A: Si, questo e' possibile e supportato.

Q from Mauro: Volevo avere qualche informazione in più sulla gestione del load balancing in ambiente Cluster.
A: La mia risposta si riferisce all'utilizzo di MySQL Cluster, non del cluster a livello di sistema operativo. Le operazioni di load balancing vengono gestite automaticamente e trasparentemente tramite il server MySQL e i data node che fanno parte del Cluster. Il bilanciamento tra il sistema client e il server MySQL deve essere gestito tramite le normali regole di balancing tra server, ovvero con bilanciatori SW, bilanciatori HW o tramite l'utilizzo delle funzionalita' di alcuni connettori, come il connettore JDBC.

Q from Eleonora: C'è un numero minimo di data node e server node nella configurazione cluster?
A: Si, il numero minimo e' di due nodi mysql, due data node e un management node, dislocati su un minimo di tre server. Questa configurazione evita qualsiasi Single Point Of Failure. E' possibile e consigliabile duplicare anche il management node, sebbene l'utilizzo di questo componente avvenga solo in fase di startup o di guasto di un altro nodo, per cui non viene considerata "configurazione minima".

Q from Stefano: master-master: l'automatic failback + multimaster autoincrement funziona anche sotto windows?
A: La replicazione funziona allo stesso modo in Windows come in Linux, Solaris o altri sistemi. Ricordo pero' che se si utilizza la replicazione per l'alta disponibilita' non e' possibile utilizzarla anche per separare le scritture con il meccanismo dell'autoincrement.

Q from Gabriele: Quando e' meglio utilizzare repliche gestite su mysql e quando e' meglio utilizzare repliche gestite dalle tecnologie di replica dello storage?
A: La replicazione di MySQL e' piu' semplice da implementare ed e' multipiattaforma, ma esiste sempre la possibilita', seppure remota, di perdita dei dati. Cio' e' dovuto alla natura asincrona della replicazione. Una replicazione con DRBD e' invece sincrona, pertanto garantisce che i dati presenti su ambedue le repliche siano consistenti.

Performance and Tuning Webinar 22 April 2008 - New Material

2008-05-01T06:13:00.009+01:00

The EMEA webinar on Performance and Tuning last week (22 Apr 2008) was a great success! Great attendance, lots of questions and an amazing job from Henrik Ingo and Erik Granstrom!

I have posted the recording of the webinar here and the slides here - unfortunately we could not post the material on our MySQL website, since it's pretty crowded now with P&T docs.

Here is the Q&A session:

Q from Ketil: I have a DB that is "read many, write few", and will remain relatively small (less than RAM). How do I tune a InnoDB MySQL-database for optimal query performance? There are 8 hasAndBelongsToMany relations to categories and a few 100 000 objects.
A: First of all, have a look here: http://dev.mysql.com/doc/refman/5.1/en/innodb-tuning.html
You should work with some system params, for example the innodb_buffer_pool_size, to set a specific amount of RAM as buffer for InnoDB.

Q from Alon: Is the query cache only available in version >=5?
A: No, it is available from 4.0.1

Q from Prakash: Is it efficient to put numerical criteria in quotes ? (eg cust_id = '675') --- [Oracle queries would be affected by this]
A: If the column you are using is a character columns, then yes, that is the way to use the index. If It's a numeric column, then you should avoid it.

Q from Konstantinos: Is there a possibility to save a whole table in some kind of cache?
A: Yes, you can use the MEMORY storage engine to keep the data in memory. Another option is to use the caching mechanisms provided by the storage engine, the key buffer for MyISAM and the buffer pool for InnoDB. At startup you can use a LOAD INDEX INTO CACHE command for MyISAM or a simple SELECT for InnoDB to cache the data.

Q from Erik: "queries must be identical" means logically or really literally ? ex.: select a.col from a is equal or not to select col from a;
A: They must be the same, byte by byte

Q from Laszlo: Can MySQL Enterprise Monitor advisor help us to determine the cache size/options?
A: Yes, it provides recommendations on how big the cache should be, if it's over or under utilised.

Q from Ketil: Are there any speed (or stability) issues I should be aware of using the .NET connector? And whichh version is the latest stable version? 5.1.5?
A: Yes, you may use 5.1.5. Starting from 5.0.X, the connector includes support up to MySQL 5.1 and ADO.Net 2.0 insterfaces and subclasses. 5.1.5 contains new membership/role provider, Compact Framework 2.0, a new stored procedure parser and improvements to GetSchema. Connector/NET 5.1 also includes the Visual Studio Plugin as a standard installable component.

Q from Saravanan: Does query cache cache sql or data?
A: It caches the result (data) of a SELECT query. The exact text of the query is used as key.

Q from Ivan: Does EXPLAIN work in other databases (apart from MySQL)?
A: Yes, EXPLAIN is an ANSI SQL Command

Q from Alfred: In ORACLE you can use variables as plpaceholders in the where criteria. Doing so, the cache is hit much more often when only some string in the criteria has changed. is this possible in mysql too?
A: it is definitely possible. The use of variables and placeholders will result in a query that can be executed using the query cache, or the storage engine cache or simply retrieving data on disk - it simply depends on the availability of the data in one of the cache.

Q from Jörgen: In which version was EXPLAIN EXTENDED introduced in MySQL?
A: It's certainly available in 4.1, I do not remember if it was available in 4.0

Q from Pierrick: Join type index is bad ? How so ? I thought it was the best join, the fastest.
A: Join type "index" means that the server has to perform a sequential scan on the index tree. Only ALL, which is a full table scan is worse.

Q from Elinor: what exactly does qcache_lowmem_prunes represent? (e.g. 3937244) and how can you be sure your cache size is right?
A: It counts the number of queries that have been removed from the cache to free up memory for caching new queries. See here http://dev.mysql.com/doc/refman/5.0/en/query-cache-status-and-maintenance.html
In other words, a low number is typically good, a high number means you should consider more memory for the cache.

Q from Prakash: In Oracle we can use hints to influence teh optimiser to take a specific path to obtain the data; Is there a similar facility in MySql ? (eg. use a specific index for a specified table)
A: MySQL has something similar, you can force the optimiser to use or not use an index. See here http://dev.mysql.com/doc/refman/5.0/en/index-hints.html

Q from Kjeld: Does the query cache exists in RAM or on Disk?
A: The query cache is in RAM, it resides within the data context of the mysqld process.

Q from Renaud: Is the query cache available by instance or by user ? I mean that if 2 users execute the same query but with not the same rights... when the second user execute the query, it will re-parse the query to check table rights ?
A: Yes, the rights for the user are checked before retrieving the data from the query.

Q from Aron: Is there a fast way of filling a memory table, because if it takes 1 minute to fill, it will lock all reads to it
A: Not really. Typical use case is to fill it on server startup and then it is only read after that.

Q from Markus: I have always "filesort" in the Extra field when checking my queries with EXPLAIN. I heard this is not the best but I do allways have this when I am using ORDER BY in the query. Is it right?
A: it means that the server must execute a sort of the resultset before providing the result to the client. This can be avoided if the ORDER BY of the query is related to an index.

Q from Hendrik: Is EXPLAIN also available in 4.1?
A: Yes, absolutely

Q from Madhusudan: I have a ndb cluster with 4 data nodes. All the 4 data nodes sit on disk-less blades (i.e. they have only RAM). Now if I create a table with storage engine as myisam or innodb, then where it will be created?
A: It will be created on the SQL node you are connected to. ndb data nodes are completely unaffected, and what's more, other mysql nodes are also unaffected.

Q from Andrea: I wrote a few stored functions in 5.1 and I noticed that the performance is a much slower than executing calculation directly in the query. Now I noticed that even the cache size commands cannot be used with functions. Any suggestions?
A: It's true that you cannot use the query cache for SELECTs with functions. Regarding the performance of your functions, I am not sure the issue is related to the version of the server. We should check which kind of calculation you execute within the function. It's anyway true that if you can execute a formula directly within a SELECT without calling a function it should be faster.

Q from Jörgen: Is there any performance difference creating a table with a PRIMARY KEY, compared to just creating a UNIQUE INDEX with the same columns as the primary key instead?
A: It depends on the storage engine and on the type of index - Btree, clustered, hash etc.

Q from Pierrick: Where can I find info re the MEMORY Storage Engine?
A: Check the ref manual, here http://dev.mysql.com/doc/refman/5.0/en/memory-storage-engine.html

Q from Ketil: What is the best practice for maintaining a sum/average? Let's say, for a given object, I want to store how many times it has been viewed. Just incrementing with select/update is not "safe", adding each entry to a table and using sum() can be slow.
A: If you just want the sum (or avg) and not at all the individual result, you can do it with one update and no select at all. This is safe. Something like UPDATE tablename SET sumcolumn=sumcolumn+1 WHERE id=...

Q from Donald: How does the size of a query influence how many queries per second can be issued? e.g. If I update blobs of 500K what would be the max QPS?
A: The bigger the query, the more you have to read and write from/to disk. This will affect the number of queries per second.

Q from Walter: Is there any performance loss when you create a view from 'large' queries?
A: A view always generates an overhead, but this is often insignificant and there is a great advantage in using views.

Q from Matthijs: Is it possible to partition a table if the data is already in it? And if so, is it also possible to change the partitioning. And is it also possible to drop partitioning?
A: Yes, all of this is possible using the ALTER TABLE command in 5.1

Q from Pascal: So, vertical partitioning is a good idea for example in a users-table, small partition with the encrypted password and id, all user data in another partition?
A: Yes, that might work pretty well. If you have a Users table with a user profile, you do not want to have a single table with tens of columns

Q from Thijs: From which version of MySQL partitioningis supported?
A: It's supported from 5.1

Q from Ricardo: What is COALESCE? What could i do with that instruction?
A: COALESCE PARTITION reduces the number of partitions when the partitioning mechanism is hash or key.

Q from John: What is the column limit for a mysql table?
A: If you refer to any datatype in general, it's 4GB for a BLOB

Q from Shane: Are all these performance advisors available at all levels of MySQL Enterprise? (e.g. basic)
A: No, performance advisors are available only with the Platinum edition.

Q from Ketil: Can I replicate an InnoDB to MEMORY, so that I can write with transactional security and foreign keys, and read really fast?
A: MEMORY is not a transactional storage engine and an in memory structure, unless it's not redundant and synchronised, is not safe.

Q from Jelle: Is it possible to get the analysuing tools without buying the enterprise server?
A: Unfortunately the performance advisors are part of the MySQL Enterprise subscription.

Q from Rami: what is the performance impact of the Enterprise Monitor?
A: the agent running on the database server needs a very small amount of memory.

Q from Dainius: is it really better to make more little sqls than one big sql?
A: It depends on the query, but in general the optimiser will work better on small queries. Also, a large query will contain a considerable number of joins and this would affect performance.

Q from J.M.: What about query strategies?. Join vs. sub-queries...
A: Joins are usually preferable to sub-queries because of the optimisation of subqueries in MySQL. But there may be some situations where subqueries would be another way to break a single query in smaller queries.

Q from John: What is the maximum number of columns for a mysql table?
A: It depends on the storage engine. InnoDB, for example, cannot contain more than 1,000 columns, MyISAM does not have this limit.

Sun and MySQL at the Royal Geographic Society

2008-04-28T13:35:00.003+01:00

Definitely a great event took place last Friday, 25th of April, in London, at the Royal Geographical Society - see http://uk.sun.com/sunnews/events/2008/apr/mysql/

My presentation concerned the commercial offerings of MySQL, covering the server, 3rd party storage engines, the tools and the OEM solutions. If you are interested you can download the slides in PDF format from here.

-ivan

Q&A and material for the webinar: MySQL Enterprise Performance and Benchmarking

2008-04-27T00:30:00.004+01:00

This is a shortlist of the questions raised during the webinar MySQL Enterprise Performance and Benchmarking that I presented the 9th of April.

The slides are available here.
The full recording of the webinar is available here.

Questions and Answers:

Q from Chris: Are you able to monitor all versions (non-Enterprise and 4.1) with MySQL enterprise monitor ?
A: Yes, you can, but some advisors will not work with versions prior to 5.0.

Q from Chris: If for example I have a couple of Enterprise systems but also some non-enterprise systems (we will be migrating over time) then would we be able to use the monitor (which we will have access to with the enterprise license) to monitor our legacy servers?
A: Yes, you can certainly do that. You can purchase the subscriptions, but you do not need to install or migrate to the certified binaries.

Q from Zenon: Where can I find details of Enterpise Subscription costs?
A: All the costs and the information are here: https://shop.mysql.com/enterprise/

Q from Douglas: if you buy the platinum licence for one server which comes with the performance advisor, will it monitor a servers performance which only has the gold licence?
A: You will be able to collect all the information, but the advisors associated to the platinum subscription will not work.

Q from Wojtek: Can I know the database usage per database user?
A: Not with the standard advisors. You need need to create your own advisor to present this information.

Q from Ole: Will we see any improvements to the optimizer in 5.1? Especially in handling subqueries and views?
A: With 5.1 we have improvements in the efficiency of the fulltext indexing. You have to wait 6.0 to see even more improvements, for example with subqueries

Q from Zenon: Is query cache just for static data? How does it handle data that has changed between the same query calls?
A: Query cache is invalidated by any write operation applied to one of the tables that are part of the stored query. It is very efficient if data don't change very often.

Q from Ronny: Novell ships the Enterprise license of mySQL with NetWare, can I monitor it with a platinum license to get the advisor?
A: Unfortunately Netware is not one of the supported platforms - there isn't an agent available for Netware.

Q from Rotem: I really want to use partitions - When will 5.1 become GA? Is it considered safe to use it now before it becomes GA?
A: 5.1 is now release candidate, we plan to deliver it GA by the end of Q2 2008. There are still some issues associated to partitioning that may or may not affect you. You should check the list of issues and bugs here: http://bugs.mysql.com

Q from Eleonora: Hi, I have a doubt about the MEMORY storage engine. Is it able to gaurantee no loss of data? What happens if the server dies before writing data to disk?
A: the MEMORY storage engine does not write any data to disk, therefore if you shutdown the server of in case of fault of the mysql daemon, the data in the MEMORY tables is lost. You may setup a backup to disk using a scheduled job, but this will not prevent some loss of data.

Q from Michael: Is it possible to tune parameters for different connections to the same MySQL server?
A: Yes, there are parameters that are session based, therefore they will affect only a specific session.

Which MyXML?

2008-04-18T21:26:00.009+01:00

Ok. We may all (maybe not really "all") agree that XML in a relational DB is a cool thing. But which kind of approach should we use to store non tabular data in a typical tabular infrastructure?

Personally I don't like hybrids. The core of a RDBMS is designed to be efficient and with a relatively small footprint. Inflating the core with some features that would probably be helpful to not more than the 30% of the projects (to be optimistic, but this percentage is a good reason to think about XML in a DB), does not justify the extra overhead.
Unfortunately, an XML storage engine. On the contrary, it would make things worse, because we would translate queries twice - Xquery to SQL then to XQuery again (the big fans of MySQL internals will forgive me for this simplification).

My interest for a XOR (XML Over Relational) approach goes back to several years ago, mainly because in my past life I have designed a couple of products that used XML quite heavily: a CMR system (Customer Managed Relationship - a permission-based CRM engine with multichannel capabilities) for a company called Bluparc and a CRM Solution Suite for Athoa Ltd ( http://www.athoa.com/...). Both environments heavily relied on Java and used MySQL as RDBMS, but not as XML repository. The adoption of two data managers instead of one has always created issues for our customers. Some of our customers used Oracle, but the XML repository was not in the DB. Why? Because (A) the MySQL at that time did not provide any XML feature at all and (B) because despite the widespread marketing message, the Oracle implementation of XOR did not fit the product at all.

Back to the original topic, there are three basic approaches to XOR (I made up the names, so forgive me if I may somewhat confuse you): XML Shredding, XML Object mapping and XML Indexing

XML Shredding
With XML Shredding every set of XML elements is converted into a table. Depending on the implementation, the columns of the RDBMS table can be associated to XML attibutes or to simple elements. This approach has been adopted by Oracle to implement XML (although the fact of being an object-relational database has helped a bit in defining a specific XML type and in handling complex elements within a set of basic elements).
XML shredding is good for query performance when the XML model is not too complex, ie when a query against the RDBMS is not translated into an endless set of joins with tens of tables. The shredding approach can also facilitate the indexing of the elements and the conversion of XML queries into SQL queries and vice versa. On the other hand, the generation of the SQL queries and the ERD can be very complex.

XML Object Mapping
This is the approach currently adopted by MySQL. Each document or element within a document is stored into a cell of a database column table. The granularity of the element usually depends on the application.
The object mapping approach is very efficient to retrieve deep hierarchies, but on the other hand it provides limited indexing capabilities and document manipulation. For example, in order to retrieve a single element in a XML document that is stored in a large TEXT column, the server has to retrieve and analyse the whole document.

XML Indexing
This approach is based on a set of predefined tables that are used to index and store the XML basic elements. Complex elements are split into several rows, named and indexed.
This approach is based on a fairly simple ER schema (although some complexity may be added to improve performance and capabilities) and it is used for "general purpose XOR". Being generic, this approach works pretty well for basic XML document (ie with no deep hierarchies). In case of complex documents, the generation of the SQL used to retrieve the document becomes a real bottleneck and increases in complexity pretty quickly.

Which Approach?
Unfortunately, a single approach does not fit all the needs. A good XOR strategy should provide all the approaches and user should decide which one would fit their projects. What's important is that the XML connectors and the XOR layer will make the physical storage and the ER implementation completely transparent to the application.

Is MySQL ready for one of these approaches?
MySQL can easily implement XML Object Mapping. In fact, object mapping is already available (in its basic form) in 5.1. What we need now is a mapping layer and a connector.

Anyone? :)

Webinar: Performance Tuning Best Practice

2008-04-18T19:08:00.002+01:00

Tue 22 April at 9am UK, 10am Central European time, we will present probably the most popular webinar on MySQL Performance and Tuning.

We will have some real hands on examples with several tips and tricks - all that we can cover in an hour!

This is the agenda:
- Performance and Tuning Tools
- Query performance analysis and improvements
- Modelling tips
- Engine specific recommendations

Webinar: MySQL for Online Applications

2008-03-25T06:33:00.001Z

Wed 26 March at 2pm GMT/3pm CET we will present a new webinar on MySQL for Online Applications. Details and registration are here.

MySQL finds its sweet spot in online and web 2.0 application, because of the scalability reliability and the ease of use of the server. In a typical approach, users start with a simple architecture, with one server, then they need to scale in order to support the increasing number of users and the demand for more computer resources.

During the presentation, we will talk about the common ways to scale-out, how to provide high availability and about the common components of a web 2.0 solution and its interaction with the database.

The Q&A will be posted on this blog.

Meetup Group and Thoughts on MySQL Proxy

2008-03-10T09:12:00.019Z

In my opinion, the last MySQL Meetup group in London was a really good event. We met Marten Mickos, former CEO of MySQL and now Senior Vice President of Sun, David Axmark, founder of MySQL and Simon Phipps, Chief Open Source Officer at Sun.

The main questions to the guests were pretty expected: is it going to change anything in the licensing, the deployment and the platforms of MySQL? The single answer is a "No, nothing is going to change". No change in the licensing (Simon Phipps has clearly stated it), no changes in the Community and Enterprise editions (at least not in the near future) and no changes in the development for Linux, Windows, Solaris and other OSes. There will be more opportunities, instead, of spreading the database in environments that are not the common target for MySQL at the moment. We are definitely not going to inflate the server with a large set of barely used features: performance, reliability and ease of use are still paramount.

In the light of this reassuring message, a new project, currently in alpha version, becomes more and more important. MySQL Proxy, the Columbus egg created by Jan Knesche, is quickly becoming a very important component of the scale-out approach of MySQL.

In short terms, MySQL Proxy is a middle layer that acts as MySQL: it accepts connections and queries from the client and it redirects them to one or more MySQL servers. In the middle, MySQL Proxy can intercept client requests and server responses, such as connetcions, queries and resultsets and it can process them using some LUA scripting. More details are available here.

I have been a big fan of MySQL Proxy since the very beginning. It is simple and powerful at the same time and looks to me like the perfect add on to a MySQL scale-out approach.
You can use Proxy in many situations: to log user requests, to improve complex security rules, for high availability or for load balancing. At the moment, the project is still at an early stage, so you should expect some issue that will be fixed; the Proxy architecture has also been redesigned in its new version 0.6 (we are at version 0.6.1 at the moment).

MySQL Proxy was the main technical topic for the Meetup group last Wednesday. We have spent about an hour reviewing some technical aspects and picturing some possible uses for it.

I have created a shortlist of what the Enterprise customers like to see with Proxy. Jan and the engineering team are working at some of these features, but it would be really great to see the contribution from the Community as well.

Flexible authentication and dynamic server connections

This is mostly interesting for ISPs and for organisations with a large number of DB users. Let's assume, for example that we need to manage thousands of users; each user has its own schema, located on a specific MySQL server.
Without MySQL Proxy, each user has to refer to a specific server and address in its applications and any sysadm or dbadm change would create disruptions.
With MySQL Proxy, the users may connect to the Proxy (and its address will not change), then Proxy redirects the connection to one of the servers, depending on a shared list with a match of user-server. The great advantage is to keep the complexity and the administration of the server side completely separated from the client side.

Proxy Failover

High Availability is nowadays mandatory for many applications and services. MySQL Proxy can achieve HA in two different ways. The first and simplest way is to connect and relate MySQL Proxy 1:1 with the client (the web/application server, for example), typically co-located on the same box. In this way, if a fault happens on the server, both the client/web server and Proxy are taken off the list of available servers.
The second and more complex approach is to use a fault tolerant layer between two or more Proxies serving multiple clients. This approach is usually necessary for 2-layer client/server applications (i.e. no web based) and when the use of a concentrated Proxy can give more advantages (for example for caching or connection pooling).
This implementation is already possible using a typical active/passive cluster (or using some specific features of the JDBC connector), but the connections will be completely dropped in case of fault of the active Proxy.

DB Server Failover

DB Server Failover can be achieved using replication between the master and the salve server, but replication is asynchronous at the moment and any synchronous replication set in this way is likely to have a negative impact on performance.
MySQL Proxy in this particular environment might have the great advantage of performing write operations on all the servers in parallel. Depending on the isolation level, it might be even possible to load balance write operation on the servers.
The main complexity of this approach is related to the implementation of a two phase commit protocol between proxy and the MySQL servers. The XA protocol might be used, but again there may be some performance issues.

Replication Balancing

There is a R/W splitting script within the MySQL Proxy package, since this is one of the most common requests from users who use MySQL Replication.
The R/W splitting will be certainly improved, but it would be really good to have the same load balancing strategy available for the JDBC Connector: random or bestResponseTime.

Sharding Balancing

Although sharding balancing can't be easily applied to all the applications, it is the real differentiator and the most intriguing approach to the scale-out philosophy available for MySQL.
Users might be able to scale-out their database solution not only for read operations but also for write operations, accessing a specific master servers. The approach should be consistent with the use of a specific "sharding key", similar to the partition pruning in the optimizer for MySQL 5.1.
Sharding balancing implies a complex analysis of the client query, but the results of a good sharding approach would be phenomenal. A sharding approach built within the application is still preferable (it would be application specific, but faster), but keeping sharding rules at Proxy level would open many possible implementations.

Proxy Hints

This feature is complementary to the automatic balancing for shards.
Using hints, MySQL Proxy could bypass the analysis of the whole query and it would pass the query directly to one or more "sharded" servers. For example, the hint would tell proxy to direct the query to the master server X, regardless the content of the query. Even more interesting, another hint would instruct proxy to send the query to each master or slave and in case of a returned resultset, Proxy would consolidate all the rows into a single (distinct or duplicated) collection.

Query Logging and Profiling

Query logging and profiling is one of the most interesting use of MySQL Proxy. The logging part is relatively simple and it can be achieved using simple rules to intercept the query and the resultset. The profiling part is more complex and requires the collection of further information from the server (we need to add more profiling features at the server level).
I am sure users will be happy to have a logging mechanism with Proxy, but the profiling part will be the cherry on the cake.

Query Caching

Last but not least, Web and Enterprise 2.0 applications would take a great advantage from the integration of MySQL Proxy with a caching mechanism. Again, this is something very dependent on the application and on the infrastructure in place, mainly due to the efficiency of the caching mechanism and to the (in)validation of the cache. Nevertheless, I can see a great advantage in caching certain queries in order to reduce the workload against the DB server.

Is it time for MyXML?

2008-02-27T01:09:00.001Z

Yes, that's correct, it's not a typo. Who needs XML over RDBMS?
So far, there has been a relatively small group of users who were demanding more XML features within or combined with MySQL. Perhaps it's because XML is not a big deal in the classic LAMP implementation of web-based services, but the more we penetrate the Enterprise market, the more XML is becoming an interesting topic. I am quite sure that we will move from the classic "nice to have" to the typical "commodity feature" in quite a short time, especially if we will improve our presence as the back end DB for J2EE apps and web services.

Let's be honest, the new nice XML features in 5.1 are still away from a good and proper XML coverage. They are great if we are going to store XML elements within columns of a table, but they are not really helpful if we need to use an XML interface to interact with our DB.

For what it's worth, although the fat 4 (Oracle, DB/2, SQLServer and Sybase) have more improved XML features, they are all far from being a good, general purpose solution. First of all, the XML features in this RDBMSs are all non-standard. Second, they all provide different approaches to XOR (XML over RDBMS), but none of them can address all the needs for a full and fast XML interaction with the DB.

This is really an interesting topic that we have just started to approach, and there is much more to come - Stay tuned!

The London Meetup Group and my wishes for MySQL Proxy

2008-02-27T01:06:00.003Z

There is a big event for the London Meetup Group coming up next week. We will all meet on Wednesday 5 March @ 7pm
The location: The Lamb
4 Lambs Conduit St
Bloomsbury
London
WC1N 3LZ

We will have a very special guest: Marten Mickos, CEO of MySQL (now in his new role, due to the Sun acquisition). Marten will join us and he will be happy to answer to questions, to share opinions and spend some good time with MySQL Brits users in front of a nice drink (no surprise which one will be the best seller :) ).

From a technical point of you, I have prepared a short presentation on MySQL Proxy, with a basic intro on the project and with some ideas on how to use it and what to expect from a production version. The slides will be uploaded after the venue.

As usual, I am open to suggestions and topics to cover/discuss during the venue, in group or face to face. Feel free to drop me an email or to post a comment on this blog at your convenience.

Thanks a lot to Mark Baker who is working hard behind the scenes to organise this event!

Webinar - Another webinar on MySQL Enterprise

2008-02-27T00:52:00.007Z

For the MySQL users who are still unfamiliar with the new features of MySQL Enterprise (2007 Fall Edition), I will present a webinar on this topic on Monday 3rd of March @ 3pm CET. The webinar will be in Italian, but the slides are rigourously in English, so if you are interested, just wait approx one week and you will be able to download slides and recording of this event.

The latest version of MySQL Enterprise is powered with an improved version of the Enterprise Monitor, which has a new set of Replication advisors and new set of cool graphs, to instantly control your servers. There is also a big improvement in the overall scalability of the Enterprise Monitor.

During the presentation we will discuss these new features and we will see the latest improvements in the Enterprise binaries.

If you are interested in any specific detail associated to this topic, feel free to comment on this blog or to drop me an email.

To register to the webinar, follow this link.

Webinar - Reduce the TCO with MySQL

2008-02-18T23:07:00.010Z

I have presented this webinar in Italian on Mon 12th Feb. The main topic was the scale-out strategy with MySQL, based on Replication, MySQL Cluster, in combination with other HA solutions, such as DRBD. This picture shows an example of shard using DRBD and MySQL Replication.

The recording (in Italian) and the slides (in English) of the webinar are available here:

Here is an abstract of the Q&A session (still italian):

Q from Giovanni: Volendo implementare una soluzione Mysql in un ambiente completamente virtuale (macchine virtuali su piattaforma esx server clusterizzata) è preferibile aumentare semplicemente il numero di mysql server (tante macchine virtuali) oppure usare in solo server

A: Ci sono varie considerazioni da fare sugli ambienti virtuali rispetto ai server di database. In linea di principio, la virtualizzazione di un DB server non e' un concetto sbagliato, ma aggiunge un ulteriore strato che potrebbe creare difficolta' nella corretta gestione delle risorse del server. Per contro, un ambiente virtualizzato risulta in generale piu' affidabile e piu' semplice da amministrare.

Q from Paola: ci sono guide in Italiano?

A:Esistono whitepaper e case study in Italiano, scaricabili dal sito, anch'esso in Italiano, ma non ci sono al momento manuali di riferimento in Italiano. La guida per la certificazione MySQL 5 e' oggi anche in Italiano.

Q from Fulvio: E' possibile avere il DB alfa su macchina A come master e beta come slave e sulla macchina B alfa come slave e beta come master?
A: Si, questa e' una topologia supportata, seppure necessita di alcune attenzioni (in particolare legate al fatto che la replicazione non opera nessuna risoluzione di conflitti).

Q from Mario: Qualche chiarimento su mysqlproxy?A: MySQL Proxy e' un nuovo software, oggi in versione beta, che permette di monitorare, amministrare e modificare le connessioni, le chiamate e i risultati inviati/provenienti da/verso server MySQL. Maggiori informazioni sono disponibili qui.

MySQL at the MEvents in Milan - Tue 29 Jan 2008

2008-02-04T09:47:00.001Z

Good location and very good presentations from our partners Magirus and Red Hat, in Milan last week.

Giuseppe Paterno', Senior Solution Architect at Red Hat (here is Giuseppe's web site, in Italian) has delivered a great presentation on the new features of RH 5.1. I think that the new virtualization features, now included in 5.1, will attract the interest of many Red Hat and MySQL users. There has been a long debate whether DB servers should stay away from virtualization or they should get the big advantage of the availability and the flexibility that a virtualised environment can provide. I think that para-virtualization may be a good compromise between a thick virtualized layer and a non-virtualized environment: it's more predictable, less intrusive and it can provide all the flexibility required for a DB server. More information on virtualization and on the new features of RHEL Advance Platform can be found here.

Luca Gargaglione, the Country Sales Manager for MySQL in Italy, has commented the acquisition with the audience and we have gathered a very positive feedback.

The technical presentation was tightly connected to various case studies. Enterprise Unlimited (an all-you-can-it subscription at the price of a single CPU of Oracle Enterprise: € 32,000) is still going on in 2008 and more enterprises are taking advantage from is really an interesting offer. In addition to that, we have anticipated part of the roadmap of MySQL Enterprise, that make this product even more enterprise-ready.

We have had lots of questions on the storage engine architecture and a growing interest in Infobright Brighthouse, a column-based storage engine aimed at providing a specialised storage for data warehousing.

One of the attendees made an interesting comment during the Q&A session: he suggested to increase the efforts in explaining the value of the open source for the Enterprise and in clarifying the boundaries of a certified version of the software. The main concern is that the public usually listen to two main messages: open source software can be analysed and modified by the user and open source companies certify and support this software. Users should be reassured that they are able to do both, but if they need to modify the software and still need certification and support, they need to talk to the software providers.

The slides in PDF format are available on the FTP server and on a slow HTTP server.

I've almost forgot: at the end of the event, the beer tasting session was really good :) !