Ting Hsuan Lin | September 28, 2016
All cloud service providers understand the strength of Open source, and embrace it. A lot of cloud services are based on open source, especially on Big-data and container domain. Open source proved to be a very quick way for cloud provider to develop cloud capabilities and then deliver it with scale. This also brings many benefits: Good relation with base developer community, familiar to use, easy integration with existing ecosystem, and very little lock-in. All the clouds are pushing for container and beyond. AWS is leading the race with ECS and Lambda. Azure follows with Azure Linux Container Service, and Windows Container will be introduced early next year. IBM Bluemix also supports container through IBM Containers. We see increasing appearance of Microservice in design and DevOps in development/operation of cloud applications in all platforms. Many IaaS providers like HP (and Rackspace before) have retreated from the public IaaS cloud to focus on managed cloud. It is an alternative offering when you provide both infrastructure and managed services at a slightly higher cost.
Amazon Web Services
Amazon Web Services has confirmed their status as the clear leader of the IaaS market. According to Gartner, AWS has five times IaaS compute capacity in use than the total of the 14 other providers combined, including Azure. While it still has unrivalled status among developers, only recently does it focus heavily on the enterprise market. There was multiple new services in two domains: Security and Migration, the pain points that enterprise customers often cite as obstacles blocking the public cloud path. Last year there was Amazon Aurora (a MySQL database with 5x times more performance than tradition MySQL and 10x times less expensive than current enterprise database offering). This year there is Amazon QuickSight (a business analytics engine which supports all types of database and AWS storage services, with 10x times less expensive than traditional BI). Amazon Database Migration Service is a new tool that eases migration to AWS database service. On Security domain, we have AWS Web Application Firewall, Amazon Security Inspector and AWS Config Rules. With the introduction of new services, Amazon has declared its intention to go deeper into traditional enterprise software territory. Amazon Web Services also consolidated its excellent reputation among developers by introducing AWS Lambda, a new cloud computing model. With traditional models like virtual machine and container, AWS provides them through virtual machine that customers have to manage. With AWS Lambda, customers can simply define a function and when the function should be triggered (by event or schedule), and AWS Lambda will run this function for the customer. The cost is calculated according to the times the function is run. Overall, Lambda can start a new evolution in cloud computing, after virtualization and container. Finally, AWS has joined in the IoT race with the introduction of AWS IoT, a cloud-based platform that allows easier development of IoT applications and tightly integrated with existing AWS services including Kinesis, Lambda, and CloudWatch. In its analysis, Gartner “recommends AWS for most cloud IaaS production deployment scenarios”. AWS is the market share and feature leader in public cloud IaaS.
Microsoft Azure is the second best alternative behind AWS. However, Microsoft has a strong root in enterprise world, a strong ecosystem (including Windows, Office, SQL Server, and .NET), and a strong developer community that they can leverage. After Satya Nadella took the helm at Microsoft, many things have changed. Cloud has become the backbone of everything and not as a second thought like before. Office has gone to the cloud in the shape of Office 365 and thrived. The story about Azure is more complicated. It has a slower start than AWS. Initial effort to differentiate Azure against AWS has been chaotic and not very successful. That slow start means Microsoft often plays catch-up. Currently some core services (Network, Security, RBAC) of Azure still lag behind AWS in term of service feature and service maturity. Under Nadella, Azure seems to have to find the right balance:
Azure is currently the cloud with highest growth. In its annual earnings release, Microsoft confirmed that Azure has seen 100% increase in revenue. Gartner advised that it is safe to consider Azure for many projects, deployments and applications. Although Azure trails AWS in total score and in some technical offerings and configurations, many enterprise customers choose Azure for the integration with the overall Microsoft ecosystem and on-premises software.
IBM Bluemix and SoftLayer
IBM’s situation is like a large giant that have problems turning and catching up with faster competitors. Watson, an advanced Artificial Intelligence platform, the long-term bet of the company, has been slow to deliver. Bluemix and SoftLayer, while gaining customers, has been unable to offset the loss from IBM’s traditional server business. While every other competitor has one single cloud offering, IBM divide its cloud offerings into two separated items: SoftLayer (IaaS) and Bluemix (PaaS). While Bluemix is run in SoftLayer’s infrastructure, they are not closely integrated, and it is easier to consider them as separated offerings.
IBM acquired SoftLayer in 2013, and since then there was not much change. SoftLayer provides traditional IaaS service. Overall it lacks in features compared to AWS or Azure. It can be seen as a managed cloud service, which provides managed services for its customers at a slightly higher cost. Gartner does not advise IBM SoftLayer IaaS to enterprise clients looking to host production workloads in a public cloud IaaS provider without additional managed services. SoftLayer should be considered only as part of a much larger IBM outsourcing arrangement that may include data center outsourcing or other managed services. Still new to the market, Bluemix has a slow start and is a long distance behind its competitor, AWS and Azure. Some of its movements are:
Google Cloud Platform
Google Cloud Platform is a curious case. Born in 2008 as Google App Engine (GAE), it is one of the earliest cloud providers. It has good relationship with developers and an attractive free tier support. However its multiple restrictions at that time (Limited language support, no file support, does not use relational database, and require a different design mindset…) means that it was only suitable for a very limited subsets of web applications. In addition, there were multiple downtimes and performance issues. That resulted in customer often develop applications using GAE’s generous free tier, but then go to Heroku or AWS to host the production version. Recognizing that GAE does not live to its full potential, Google chose to rebuild its cloud offering as Google Cloud Platform, which contains the old Google App Engine, the newer Google Compute Engine and Google Container Engine, and multiple proprietary storage and big data services. While its offering is still not as comprehensive as its competitors like AWS and Azure, Google is looking to build up Google Cloud Platform by introducing much needed services like Cloud Monitoring, Cloud Deployment Manager and Cloud DataLab. It is also learning from its competitors by providing several similar services and features: Preemptible VMs (Spot instances), Nearline (Glacier)… Gartner finds that Google Cloud Platform is missing many features for production-grade enterprise workloads, lacks differentiation and doesn’t support hybrid integration as well as its competitors. Gartner recommends only limited use of Google Cloud Platform for cloud IaaS production deployment in scenarios where its shortcomings are not problematic.
Actions of FPT
From the previous year until now, FPT has been developing Cloud Professional Services offering and products and bringing the whole services to customers from consultancy to implementation and post-implementation. The solution enrichments are Citus Cloud Suite that was also mentioned in this magazine. In 2015, these services and product will be able to improve with the new features and up to new versions. The Cloud team focuses on containerization and Docker, makes it become a framework for some projects.
AUTHOR – VietPH –