As the flag-bearer of open source software, Red Hat has seen its fortunes grow as more companies turn to Linux and, more recently, containers and microservices to power their businesses.
During its fourth quarter of the 2017 fiscal year, Red Hat raked in revenues of $629m, up 16% year on year. Asia-Pacific was the fastest growing region for the company, contributing 14% of total revenues for the year.
In a wide-ranging interview with Computer Weekly, Damien Wong, vice-president and general manager for Red Hat in ASEAN, sheds light on the company’s success, its strategy for tackling the Southeast Asian market that is not used to paying for software, and the role of open source in digital transformation.
Red Hat is widely considered one of the world’s most successful open source companies. What do you think the company got right from the start?
Wong: For a 24-year-old company, our business model has been amazingly consistent. We started off with an open source subscription model, which meant we had to work harder to earn our business and strive to remain relevant to our customers.
As an open source company, we don’t have proprietary software, unlike some purported open source companies that have an open core with proprietary add-ons to make their offerings enterprise-ready. That doesn’t make them open source companies in the strictest sense. Companies that adopt such software are still beholden to proprietary components.
Neither are we an altruistic company that will only support community versions of open source software, because that does not instil confidence among enterprises. While the open source community is great at delivering innovations and new features at a rapid pace, not everyone is good at maintaining reliable, stable and secure software – the not-so-glamorous aspects of software that enterprises need. While we need to leverage the innovations from the community, trying to support the community version is extremely difficult – if not impossible – because of the rapid evolution of technology.
So what we’ve done is to provide stable versions of open source software, not only by hardening it, but also making sure it is compatible with different applications and hardware. This enables enterprises to continue using our software, knowing that it will be stable, secure, and will perform well. Our model does not go against the principles of open source, because everything we do goes back to the open source community. But at the same time, we’re not naïve enough to say we will support the latest and greatest community versions of open source software.
During the recent Red Hat Summit, executives acknowledged that open source tends to work better at the infrastructure layer, rather than in applications. Why do you think this is the case?
Wong: It’s true that open source software is more mature in infrastructure than in business applications. I think it’s a case of having a bigger critical mass of users in the infrastructure layer (which is fairly standardised and commoditised), where open source has been proven to deliver a competitive edge in a number of use cases within a snapshot in time. This has made it easier for enterprises to adopt open source software, thrusting it into the mainstream. That said, I think it’s not a case of open source not being applicable to business applications, which tend be more customised to the needs of a specific organisation or industry. It’s just a matter of time before those customisations become open source, which is the preferred mode of innovation today.
Do you think the emergence of microservices will speed up open source developments in business applications, by getting enterprises to think about interoperable open platforms when deploying applications, and not just in the infrastructure layer?
Wong: For sure, one of the key concepts around microservices is reusability. So if you create a microservice and abstract it well enough, it can be used across different applications. I think the evolution of microservices and their maturity in the application ecosystem will lead to a situation where you can pull together microservices from marketplaces in an application that serves a purpose well. When that will happen depends on market demand, and when there’s a strong need, the open source community will come together to address that need.
Selling free software can be hard, especially in ASEAN, where a majority of people and businesses are not used to paying for software. Although Red Hat is growing at healthy rates in the region, do you see this as a stumbling block for future growth?
Wong: Southeast Asia has varying levels of economic growth and maturity in technology adoption. Open source provides organisations with access to technology that they otherwise could not afford. While not all organisations will survive, those that do will find that they need security and performance from their software, like any large enterprise would. Take Grab, the ride-hailing company, for example. Its uptime of 87% when it first started out had impacted the livelihoods of its drivers. They worked with Red Hat and made use of Ansible to automate the roll-out of application changes, increasing availability to more than 99%.
Traditional enterprises, on the other hand, are used to paying for software, though this is not the case across the board. In some emerging countries, even large organisations have misperceptions about open source. Some may still be using community versions of open source software without enterprise subscriptions, which we don’t encourage as they roll out mission-critical applications.
Damien Wong, Red Hat
When deploying mission-critical applications, it is no longer just about creating a sandbox to test out new concepts – it will affect the customer experience and financial transactions. If a patch for a known vulnerability is available and you don’t have access to that patch because you’re on a community version, you’ll face real issues. While we make security patches available to the open source community, it may take weeks before they make their way into community versions because of open source governance processes. If you’re a bank, you could potentially expose customers to unnecessary risk and liability.
Why do you think large organisations in ASEAN still use the community versions despite the risks you have just described? It just doesn’t seem rational. Is this about saving money, thinking that they can fix any problems on their own?
Wong: We can only guess why they do it. Some companies may hire good technical people, thinking that it’s a technology risk rather than a business risk. Having a very smart engineer using community software to support a production system may not be an issue. He can check with the forums or download the patches, so it’s just a technology thing. The business risks that I talked about may not be so apparent to an engineer or developer. We will have to educate them that open source is not just the domain of IT – it’s the domain of the business as well. With software being seen in some countries as something you download from the internet and not what you pay for, it will take some time before software is seen as something that has to be taken seriously.
In recent years, Red Hat has been touting the role of open source in digital transformation, at least in this region. With almost every other technology company spreading the same message, how is Red Hat making itself heard?
Wong: Many people may not realise it, but a lot of innovation such as big data and cloud arose from the open source community. These technologies gave rise to cloud-native, digital disrupters, which have disrupted nearly every industry, from transportation and retail to hospitality. The traditional companies that were being disrupted then started to look at what they needed to do to stay relevant to their customers, and that’s when the concept of digital transformation became really popular. So the root of digital transformation has been open source innovation.
Of course, there have been proprietary companies that have tried to emulate open source innovation. But time and time again, we’ve seen how open source communities have always out-innovated those companies. And Red Hat, being a proponent of open source, is participating in many open source projects that are driving digital transformation. For example, we talk about DevOps because every company needs to develop applications in a faster, more agile manner. That means moving away from monolithic infrastructure to things like microservices and containers, which Red Hat is backing with its contributions to the Docker and Kubernetes projects. As a leader in the open source world, it is natural that we take leadership in digital transformation as well.
Read more about IT in ASEAN
- Malaysia’s Cyberjaya township will pilot a slew of smart city projects, including e-payments and mobile bus ticketing.
- The Singapore government will dish out an estimated S$2.4bn worth of technology contracts in the 2017 financial year to develop a national sensor platform, among other initiatives to speed up the country’s digitisation efforts.
- Nearly 9,000 command-and-control servers and hundreds of compromised websites have been identified across the ASEAN region in a cyber crime operation led by Interpol.
- IT head honchos in ASEAN will spend 2.2% more in 2017, mostly on digital initiatives aimed at transforming their businesses.
Open source projects like Kubernetes were started by companies like Google, one of the biggest users of open source software. Why do you think that is the case? I would think Red Hat, as an open source leader, would be the one starting those projects.
Wong: That’s a good question that captures the success of Red Hat. We don’t espouse the belief that if a piece of open source technology is not invented here at Red Hat, it’s not good. In fact, we actively look at projects developed by others that might be superior. A good example is OpenShift, which is now fundamentally made up of Docker and Kubernetes, as opposed to the original technologies that we had started with. Our customers don’t have to be afraid of being stuck with a technology that may be at a dead end, without broad community support and stubbornly backed by only one company. It’s the same case with OpenStack, which was created by Nasa and Rackspace. But today, Red Hat is the largest OpenStack contributor.
Red Hat is perceived to be using Linux to cross-sell OpenStack. Is that the strategy Red Hat is pursuing?
Wong: All our technologies are predicated on Red Hat Enterprise Linux, the foundation on which you can build stable infrastructure platforms. Like a building, if your foundation is not stable, it is impossible to build anything substantial on top of it. That is why we’ve always stood by the position that you have to build your infrastructure on something that we know is stable. If it’s quicksand, or something that can’t withstand scrutiny, we can’t back it. So without a stable base in OpenStack, which has many related components, it will be difficult for us to back that project and guarantee that things will work properly. There had been situations where organisations faced challenges because they did not understand how critical the foundation layer was going to be when they rolled out OpenStack.
For now, OpenStack’s main adopters are telcos, internet service providers such as MyRepublic and cloud service providers. Do you see other sectors benefiting from OpenStack as well?
Wong: That’s a good observation. Service providers are naturally looking at OpenStack because of the move towards network function virtualisation (NFV). The standards body that governs the NFV movement is the European Telecommunications Standards Institute, which has recommended OpenStack as the underlying infrastructure layer for NFV. This has led internet service providers such as MyRepublic to look at OpenStack. While OpenStack has also been deployed by institutes of higher learning, financial institutions and government agencies in the ASEAN region, you are absolutely right that the most advanced users are telcos and service providers. I hope the carrier-ready advantages of OpenStack will have knock-on benefits for enterprise data centres, because if OpenStack can support a telco service, it can definitely support enterprise applications.