Red Hat’s acquisition-fueled climb to the cloud

Red Hat is famous for its ability to focus squarely on a market and technology and build success from there, as it did with Linux. However, the company increasingly has diverged from its roots and historical laser focus on the enterprise x86 server market with Red Hat Enterprise Linux.

The overarching theme and identity of Red Hat is still open source software, but the main driver for the company clearly is now cloud computing, which is intertwined with open source.

Read the full article at LinuxInsider.

MySQL vs. NoSQL and NewSQL – survey results

Back in January we launched a survey of database users to explore the competitive dynamic between MySQL, NoSQL and NewSQL databases, and to to discover if MySQL usage is really declining – as had been indicated by the results of a prior survey.

The publication of the associated report took longer than expected, mostly because we expanded its scope to include revenue and growth estimates for the MySQL ecosystem, NoSQL and NewSQL sectors respectively, and with that report now published I am pleased to fulfil our promise to share the survey results.

We seem to be having some random embedding issues so for now the results can be found on SlideShare, adapted from the presentation given at OSBC earlier this week. For greater context, we have also included an explanation of each slide, below:

Slide 2: Provides an overview of the associated report – MySQL vs NoSQL and NewSQL 2011:2015, which is available here.

Slide 3: Explains why we launched the report. We once described as the crown jewel of the open source database world, since its focus on Web-based applications, its lightweight architecture and fast-read capabilities, and its brand differentiated it from all of the established database vendors and made for a potentially complementary acquisition. Today, the competitive situation is very different.

Slide 4: Oracle’s MySQL business faces competition from the rest of the MySQL ecosystem, as illustrated in Slide 5, many of which have emerged following Oracle’s acquisition of Sun/MySQL.

Slide 6: The emergence of these alternatives was triggered, in part, by concern about the future of MySQL. A previous 451 survey,conducted in November 2009, showed that there was real concern about the acquisition, with only 17% of MySQL users believing Oracle should be allowed to acquire MySQL.

Slide 7: The 2009 survey also showed that while 82.1% of respondents were already using MySQL, that figure was expected to drop to 72.3% by 2014. That survey was conducted amid a climate of fear, uncertainty and doubt regarding the future of MySQL, and one of the drivers for our current report was to see if that predicted decline occurred.

Slide 8: To put this in context, we asked the current survey sample (which included 205 database users) about their reaction to the acquisition. While the vast majority of MySQL users reported that they continued to use MySQL where appropriate, 5% indicated that they were more inclined to use MySQL, and 26% said they were less inclined to use MySQL. Not surprisingly the proportion of users less inclined to use MySQL was much higher amongst those abandoning MySQL than those sticking with MySQL.

Slide 9: We also asked respondents to rate Oracle’s ownership of MySQL on a range of very good to very bad. Overall, the balance tipped in favour of a negative perception of Oracle’s track record, while there was naturally a more negative perception of Oracle amongst those abandoning MySQL compared to MySQL mainstays. However, the results showed that the percentage of respondents rating the company’s performance ‘very good’ and ‘very bad’ was actually quite similar for both abandoners and mainstays. While those abandoning MySQL are more likely to have a negative perception of Oracle, it is not necessarily safe to assume that Oracle’s actions and strategy are the cause of the abandonment. Clearly there are other competitive forces at work.

Slide 10: Not least the emergence of NoSQL, as illustrated in Slide 11, and NewSQL, as illustrated in Slide 12.

Slide 13: Based on some very high profile examples of projects migrating from MySQL to NoSQL, there is a common assumption that NoSQL and NewSQL pose a direct, immediate threat to MySQL. We believe the competitive dynamic is more complex.

Slide 14: While 49% of those survey respondents abandoning MySQL planned on retaining or adopting NoSQL databases, only 12.7% said they had actually deployed NoSQL databases as a *direct replacement* for MySQL.

Slide 15: In comparison, there is much greater overlap between NewSQL and MySQL, but of a complementary nature. 33% of respondents retaining MySQL had considered, tested or deployed NewSQL database technologies, while approximately 75% of the NewSQL revenue for 2011 is from vendors that we also consider part of the MySQL ecosystem.

Slide 16: The results of our 2012 survey show that MySQL is currently the most popular database amongst our survey sample, used by 80.5% of respondents today.

Slide 17: However, it’s popularity is again expected to decline to 2014 and 2017. This indicates an accelerated decline in the use of MySQL, compared the findings of our 2009 survey. While that survey was conducted amid a climate of fear, uncertainty and doubt regarding the future of MySQL we are not aware of any specific reason why the 2012 sample, which was self-selecting, should have a disproportionately negative attitude to MySQL or Oracle.

Slide 18: MySQL’s predicted decline of 26.4 percentage points between 2012 and 2017 compares to a predicted decline of just 9.3 percentage points for Microsoft SQL Server, and only 5.9 percentage points for Oracle Database. In comparison, MariaDB, Apache Cassandra and Apache CouchDB are expected to increase in usage by 3.0 percentage points or greater between 2011 and 2017.

Slide 19: Although alternative MySQL distributions including MariaDB, Drizzle and Percona Server are expected to see increased adoption over the next five years, they are not growing at the same rate that MySQL is declining.

Slide 20: So where are those abandoning MySQL going to? Looking specifically at the 55 MySQL users who expect to abandon it by 2017 (which is admittedly a small sample, and therefore not to be considered statistically relevant) we see that PostgreSQL is the most popular database being retained or adopted over the same period, followed by Microsoft SQL Server, Oracle, MongoDB, and MariaDB.

Slide 21: This only tells part of the story, however. Just because a company is retaining Oracle Database, for example, does not necessarily mean that Oracle Database is being used as a replacement for the abandoned MySQL. We therefore also specifically asked survey respondents which databases they had considered, tested or deployed as a direct replacement for MySQL. The response from the 55 respondents planning to abandon MySQL again saw PostgreSQL, MariaDB and MongoDB as the most popular answers, followed by Apache CouchDB and Apache HBase.

Slide 22: While NoSQL database were well-represented in this list, we saw that anyone considering NoSQL considered multiple NoSQL databases. Per respondent, NoSQL databases were the least considered of all alternatives by existing MySQL users.

Slide 23: The survey results suggest that MongoDB is the most often considered, tested or deployed as a replacement or complement for MySQL, followed by Apache CouchDB, Apache HBase, Apache Cassandra/DataStax, and Redis.

Slide 24: NewSQL technologies that improve the scalability and performance of MySQL scored well, with eight of the top 10 most considered NewSQL technologies being directly complementing MySQL. Of the other two, one (Drizzle) is a derivative of MySQL, and the other (Clustrix) can also be used in a complementary manner as part of a MySQL cluster, although in the long-term is positioned as a direct alternative.

Slide 25: MariaDB is the member of the MySQL ecosystem most often considered, tested or deployed as a replacement or complement for MySQL, followed by Continuent Tungsten, Percona Server, MySQL Cluster, and Amazon RDS.

Slide 26: More than half of all MySQL users had considered, tested or deployed another relational database as a direct replacement, while over 40% had considered, tested or deployed a caching technology to complement MySQL. The memcached caching technology was the most widely-deployed of all the technologies we asked about, followed closely by PostgreSQL, which supported anecdotal evidence that a number of MySQL users are migrating to the other major open source transactional database.

Slide 27: For the record, the survey had 205 respondents. Primary job roles among respondents included: director/manager of IT infrastructure (18.0%); architect/engineer (17.6%); developer/programmer (15.6%); database/systems administrator (14.6%); consultant (14.1%); VP level or above (13.7%); analyst (3.4%); and line-of-business manager (2.9%).

Further survey analysis and perspective on the competitive dynamic between MySQL, NoSQL and NewSQL is available in the MySQL vs NoSQL and NewSQL report, which also includes market sizing and growth predictions for the three segments.

451 Research delivers market sizing estimates for NoSQL, NewSQL and MySQL ecosystem

NoSQL and NewSQL database technologies pose a long-term competitive threat to MySQL’s position as the default database for Web applications, according to a new report published by 451 Research.

The report, MySQL vs. NoSQL and NewSQL: 2011-2015, examines the competitive dynamic between MySQL and the emerging NoSQL non-relational, and NewSQL relational database technologies.

It concludes that while the current impact of NoSQL and NewSQL database technologies on MySQL is minimal, they pose a long-term competitive threat due to their adoption for new development projects. The report includes market sizing and growth estimates, with the key findings as follows:

• NoSQL software vendors generated revenue* of $20m in 2011. NoSQL software revenue is expected to rapidly grow at a CAGR of 82% to reach $215m by 2015.

• NewSQL software vendors generated revenue* of $12m in 2011 (of which $9m is also considered MySQL ecosystem revenue). NewSQL revenue is also expected to grow rapidly at a CAGR of 75% to reach $112m by 2015 (including $56m in MySQL ecosystem revenue).

• The MySQL support ecosystem generated revenue* of $171m in 2011 (including $9m from NewSQL technologies). MySQL ecosystem revenue is expected to grow at a CAGR of 40% to reach $664m by 2015 (including $56m in NewSQL revenue).

“The MySQL ecosystem is now arguably more healthy and vibrant than it has ever been, with a strong vendor committed to the core product, and a wealth of alternative and complementary products and services on offer to maintain competitive pressure on Oracle,” commented report author Matthew Aslett, research manager, data management and analytics, 451 Research.

“However, the options for MySQL users have never been greater, and there is a significant element of the MySQL user base that is ready and willing to look elsewhere for alternatives,”

As well as revenue and growth estimates, the report also includes a survey of over 200 database administrators, developers, engineers and managers. The survey findings include:

• While the majority of MySQL users continue to use MySQL where appropriate, the use of MySQL is expected to decline from 80.5% of survey respondents today to 62.4% by 2014 and just 54.1% by 2017.

• Despite the emergence of NoSQL and NewSQL database products, the most common direct replacement for MySQL among survey respondents today is PostgreSQL, which is also the focus of a recent burst of commercial activity.

• While 49% of those survey respondents abandoning MySQL planned on retaining or adopting NoSQL databases, only 12.7% of MySQL abandoners said they had actually deployed NoSQL databases as a direct replacement for MySQL.

“While there have been some high profile example of users migrating from MySQL to NoSQL database, the huge size of MySQL installed base means that these projects are comparatively rare,” commented Aslett.

The report describes how NoSQL database technologies are largely being adopted for new projects that require additional scalability, performance, relaxed consistency and agility, while NewSQL database technologies are, at this stage, largely being adopted to improve the performance and scalability of existing databases, particularly MySQL.

“NoSQL and NewSQL have not made a significant impact on the MySQL installed base at this stage but MySQL is no longer the de facto standard for new application development projects,” said Aslett. “As a result, NoSQL and NewSQL pose a significant long-term competitive threat to MySQL’s dominance.”

MySQL vs. NoSQL and NewSQL: 2011-2015 is now available to existing 451 Research subscribers. Non-clients can apply for trial access to 451 Research’s content.

*451 Research’s analysis of MySQL, NoSQL and NewSQL revenue is based on a bottom-up analysis of each participating vendor’s current revenue and growth expectations, and includes software license and subscription support revenue only. Revenue line items not included in these figures include hardware associated with the delivery of these services, revenue related to applications deployed on these databases, traditional hosting services, or systems integration performed by the vendors or other third parties.

The revenue estimates do not take into account unpaid usage of open source licensed MySQL, NoSQL and NewSQL software, and therefore represent only a fraction of the total addressable market. Based on the above revenue figures and other analysis, 451 Research estimates that the total value of the MySQL ecosystem in terms of ‘displaced’ proprietary software might equate to $1.7bn in 2011, while the NoSQL market had a displaced value of $195.7m and the NewSQL sector a displaced value of $99.4m.

Open source software and the need for speed

We’ve been spending the week at the 451 Group’s 4th Annual Client Conference and speaking to vendors, investors and end users to get their latest perspectives on what is driving open source software in the enterprise.

One consistency among all of these different groups who produce, invest in, provide and use open source software is that while the typical open source advantages of cost and flexibility are still very significant, the biggest driver at the moment appears to be speed.

I think I first began noticing the importance and prominence of development and deployment speed in mobile open source software. As discussed at our client conference panel on the topic, we see hardware manufacturers, device players, wireless carriers and others all looking to mobile Linux and open source software to respond to Apple’s iPhone, which has also served to prove that a ‘non-mobile’ vendor can quickly and effectively stake a claim in the mobile market. These organizations realize that producing their own new operating system from scratch is neither realistic nor pragmatic, given the time and investment it would take. These companies are, however, looking to mobile Linux and open source software as a way to use existing, stable software as the basis for their own branded software and services. Examples include Google and Android, Palm’s WebOS, LiMO and Symbian, which is now being open sourced by Nokia and other backers that are part of the Symbian Foundation. The fact that we have gone from bascially one single Android smartphone in the market a year ago to the cavalcade of Android devices now arriving in various forms from different vendors is indicative of the speed at which open source software can move.

Throughout my conversations with folks attending our event, I’ve heard the speed theme again in other sectors and segments. Of course, application development is a fast-moving proposition, so again, we see vendors looking to open source software as a tool that can shorten their time to market. Again, cost of development, flexibility, customization, lock-in all loom as factors in favor of open source, but the single biggest driver again comes down to speed.

We’ve also seen speed as a factor for building and providing cloud computing infrastructure. Vendors report that Linux is ideal for cloud construction since the availability of source code means that unecessary pieces can be relatively easily and quickly cut out of the OS. In addition, we continue to see a blending of roles between software developer and system administrator/operations. The rise of the devops is also indicative of the need for speed. Developers are pushing to get software released and vetted. Administrators that might otherwise resist cloud computing or other models that may cause some concern about keeping their jobs are being forced to embrace cloud computing anyway. Why? The answer, again, is speed.

Open source is obviously often viewed as a part of agile development and more effective software development, as well as distribution. As the pressure to keep up in mobile, cloud computing and elsewhere continues to build, it will be interesting to see how far open source software’s speed advantages will take it.

SpringSource-Hyperic not so open source, not so bad

There’s quite a bit of discussion about SpringSource’s annnounced acquisition of Hyperic, with all kinds of speculation on SpringSource’s competitive positioning, its desire to add a management component and, as we report, the importance of Hyperic’s focus on monitoring and managing applications in virtualized and cloud environments. What does not seem to be quite as prominent in the discussion of this deal between two companies we consider open source vendors is: open source software.

Sure, there are technical and cultural reasons that an acquisition and merging of the comnpanies makes sense. Both use primarily GPL-licensed open source software at the center of their offerings. Both offer a variety of proprietary extensions and enhancements in an open core model. Both started around the same time and are at similar places in their evolution along with the commercial adoption of open source. However, as we’ve seen in other cases when open source matters more on an underlying level, open source software is somewhat of an aside in the SpringSource-Hyperic acquisition.

That is not to say, however, that this deal reflects a lack of momentum for open source software, whether we’re talking enterprise Java and SpringSource or applications monitoring and management via Hyperic. Some are arguing that this acquisition represents SpringSource protecting its own viability by saving Hyperic, which is embedded in SpringSource’s products and business. However, given the growth we’ve seen at both companies and the continued acquisition strategy from SpringSource, that does not seem to be the case. As far as the contention enterprise and midmarket organizations are turning away from open source amid difficult economic conditions, we are seeing the exact opposite. The risk/benefit question on open source software is now falling down on the side of trying it out, and open source seems to be passing the test. Those who continue to forbid or resist open source are sounding more unreasonable, particularly when there are most likely parts of all organizations that already know the benefits, including cost-savings.

We believe the main drivers of the SpringSource-Hyperic acquisition and the main value here is not open source software, but rather where open source software has allowed both of these vendors to go: enterprise Java applications, virtualized and cloud computing environments in particular. The fact that these vendors are joining up in the latest virtualized and cloud environments indicates they have the right offering at the right time, and we expect the combined company will leverage this even more now.

LinuxWorld 2008 – nobody cares

There are certain phrases that we tend to hear a lot from vendors — ‘enterprise-class, best of breed, customer choice,’ etc. However, I was repeatedly hearing somewhat surprising phrases as I made the rounds at LinuxWorld this year: ‘We don’t care, customers don’t care, no one cares …”

Don’t get me wrong. Linux and open source have not reached the point where the software is so good, vendors and customers don’t have to care about it. The point seems to be this: there is less concern or ‘care’ about whether the operating system is Linux, Windows, Solaris or other; fewer customers care whether the software in use is open source or not; and there seems to be a general recognition that the fact a product or vendor is open source does not matter as much.

Sure, open source is still a significant differentiator. It allows vendors to get software and products into customer hands more quickly and broadly. It typically provides significant cost savings to customers. However, it is far less exotic and foreign in the enterprise, both for vendors and customers, who seem to be viewing open source not as religion, philosophy or idealism, but as just another option.

Vendors supporting various operating systems indicated there is less care about the underlying OS. Part of this can be attributed to virtualization, which allows servers and VMs running different operating systems to be managed in a unified manner. Still, even the difference between virtual and physical servers seems to be of less care to vendors, which are now moving to support and include both in their products and plans. Other vendors discussed how the use of virtual appliances and cloud computing were minimizing how much care centers on the OS, since it is becoming less visible to partners and users.

As for those users and customers, who are playing an increasingly significant role in Linux and open source, there also appears to be less care about whether software is open source. Instead, customers have come to expect comparable or superior features and functionality at less cost. Open source is often the way vendors and their products get there, but customers don’t really care.

So does all this lack of care mean that open source is in danger of losing its edge? I don’t think so. Rather, it is further testament to the continued enterprise maturation and acceptance of Linux and other open source software.