Cloud Brings out New Equations to the Ecosystem

In the era of Cloud, it is unrealistic to think of just one player providing end-to-end capabilities for the enterprises. Whether it is infrastructure services or platform services, there are a set of essential services that are expected by the enterprises. Most of the enterprise customer requirements go beyond the basic set of services like compute, storage, networking and databases. For example, a CDN service is essential to complement the storage, a NoSQL database to handle the scale out operations on the data tier and Hadoop for big data processing and analytics. Infrastructure service providers want to leverage the investments by exposing a PaaS layer. With the rapidly evolving landscape of Cloud driven by the customer needs, traditional platform companies like Microsoft, IBM and Oracle cannot match the pace. They cannot any more play the role of the big brother in the context of the Cloud. Amazon Web Services comes close to being the Cloud platform that offers a native set of building block services, application services, management services and deployment services. Cloud brought everyone to the same level playing field. Of course, an early mover like Microsoft has an advantage over IBM and Oracle but they still cannot enjoy the same dream run they had during the last decade.

Let’s look at Windows Azure. For 3 years, Microsoft tried it’s best to convince the developers that its Cloud platform is complete and can be used to deploy web scale polyglot applications. When they finally launched IaaS last year, Microsoft realized that the customer needs are beyond the core capabilities of Windows Azure. So, instead of building an email-as-a-service like the Simple Email Service (SES) of AWS, Microsoft partnered with SendGrid. Similarly, they partnered with ClearDB to bring highly available MySQL-as-a-Service to Windows Azure. During the same period, AWS has added the support for MS SQL and offered Multi-AZ capabilities. Through it’s partnership with Yahoo!, Microsoft borrowed the Hadoop stack from Hortonworks and rechristened that as Hadoop on Azure. The list goes on with MongoHQ providing MongoDB as a Service and New Relic offering application monitoring to Windows Azure developers. While Microsoft enjoyed the status of the big brother during the last decade, it is now reduced to yet another Cloud platform and relies on exactly the same 3rd party services that its competitors expose.

HP Cloud is not an exception to this phenomenon. Except the physical hardware, there is very little of HP software stack running on their Cloud. HP Cloud’s building blocks are based on OpenStack services like Nova, Swift, Keystone and Cinder. They partnered with Akamai to extend the storage to the CDN. HP’s PaaS layer is built on ActiveState Stackato which is based on Cloud Foundry. Datastax whose solution is built on Apache Cassandra and Apache Hadoop powers the Big Data offering of HP Cloud.

New IaaS players like Dell and Oracle will follow the same by assembling various services to compose their Cloud. Dell’s announcement of Fast PaaS, which is based on Cloud Foundry is an example of the same.

Like Microsoft, Google is also extending their PaaS investments to Google Compute Engine. They also have most of the building block services in place and that positions Google Cloud as one of the complete Cloud platforms.

- Janakiram MSV, Chief Editor, CloudStory.in