Database clustering specialist Continuent Inc is working with MySQL AB to provide the open source database company with a solution to Oracle Corp’s acquisition of the company behind its storage engine of choice.
Emeryville, California-based Continuent’s CEO, Eero Teerikorpi, told ComputerWire that its is working with the open source database company to come up with a solution designed to alleviate the concerns of MySQL users following Oracle’s acquisition of Innobase Oy.
Teerikorpi could not be encouraged into revealing details of how Continuent and MySQL will work together, but said an announcement is imminent, and hinted that it has more to do with people and expertise, than technology.
It has to do with the fact that our company has the best knowledge of the InnoDB product outside Innobase, he said, noting that Continuent has gained experience of the open source storage engine while building clustering technologies for the MySQL database.
Teerikorpi maintained that Oracle’s acquisition of Innobase and the InnoDB code base is not a huge threat to MySQL, but said that it was working with the company to reduce any doubts in the minds of customers.
InnoDB is an important part of the current architecture, but it’s not the only transaction engine, and it’s open source so they can keep using it, he said. There is a cloud of uncertainty [for MySQL customers]. That’s something that needs to be tackled, he added.
Asked for comment, MySQL worldwide communications director, Steve Curry, said in a statement: We are exploring several options for new MySQL database engine technology, including in-house development and speaking with several partners. As for InnoDB, Oracle has assured MySQL of its intent to negotiate an extension of the relationship – and we look forward to working with Oracle as a business partner.
Continuent has a vested interest in keeping MySQL’s customers happy. Known as Emic Networks until a change of name last week, the company provides high availability clustering technology for MySQL’s eponymous database management system with its m/cluster product.
The change of name coincided with the announcement of $5.75m in Series B funding and plans to launch versions of the */cluster technology for the open source PostrgreSQL database, as well as database products from Oracle, Sybase Inc, and Microsoft Corp.
We came to the conclusion we needed to expand, explained Teerikorpi, noting that the hiring of core developers behind the open source Sequoia project (formerly known as Clustered-JDBC) last year gave the company the ability to provide high availability clustering for multiple database products.
It solved the problem we had in the past as we don’t have to go single track, and can address a much larger market, he said. The p/cluster version for PostgreSQL will be launched early in the first quarter of 2006, while a partnership with PostgreSQL vendor EnterpriseDB Corp is also in the works.
That will be followed by the s/cluster and ms/cluster versions for Sybase and Microsoft SQL Server in the second quarter, while the o/cluster version for Oracle will arrive in the second half of the year.
The company will also examine opportunities for other databases such as Ingres and Firebird, where it makes sense, according to Teerikorpi, while IBM Corp’s DB2 also has potential. It’s not just can we create the product, it’s can we provide the support, and right now, the answer is no, said Teerikorpi of a DB2 version.
The $5.75m in funding from Trident Capital, Nordic Venture Partners, and Ledstiernan will help the company to boost its marketing, sales and product development initiatives, and bring the total venture capital funding raised by Continuent to $15.75m
Continuent is expecting significant interest from Oracle customers given that, according to Teerikorpi, only 3% of Oracle’s customers are using its Real Application Clusters technology. Most customers seem to think of clustering as a complex and expensive problem, he said.
Evidence of potential interest in the Oracle port comes from the fact that 60% of the users for the open source version of the clustering software are using Oracle and doing the integration to get the technologies to work together themselves, according to Teerikorpi.
That said, Continuent is also expecting sizeable interest from current Sybase and Microsoft database users. In the short run we expect Sybase and SQL Server to be generating revenue before Oracle because the Oracle customers we think will be more demanding, and Sybase and Microsoft do not have these sorts of technology.
While the forthcoming SQL Server 2005 will introduce data mirroring functionality Teerikorpi said he is expecting customers will have to pay a premium for the technology, creating opportunities for Continuent at the lower end of the scale.
While the */cluster technology is competitively priced at under $5,000 per database CPU, cost is not the only advantage it has, according to Teerikorpi. By supporting multiple database products, either separately or within the same environment, */cluster gives users flexibility in database purchase decisions, he said, and can also be used to enable migration projects.
We haven’t really pushed that angle yet, Teerikorpi admitted of the migration project potential. We need to get the commercial technology out there first. Once available, the commercial versions will enable upgrades (for example from Oracle 9i to 10g) and migrations (from Oracle on Solaris to Oracle on Linux, or from one database flavor to another). The problem today is you have to bring the application down, but we can offer this on the fly.
While the */cluster technology has some advantages over database vendor specific solutions, Teerikorpi is wary of talking the technology up too much. We may have a better technology, we may have a better product, but would anybody realistically believe us? We have to go out and prove the product first, he said.