By Bill Reid, Senior Director, IT Platform Strategy and Engineering
Developing Platform-as-a-Service (PaaS) is a vital part of corporate IT operations these days as such operations strive to become more responsive to their internal business customers by shedding the traditional model for IT-as-a-Service. Through the EMC Federation, we have multiple options that we have embraced to support both new and legacy capabilities, including Cloud Foundry from Pivotal which is well-positioned for the Third Platform supporting multiple Big Data and mobility needs and VMware tools supporting the automation of public, hybrid and private cloud management.
While we embrace all of these capabilities within EMC IT, we are also taking a broader, more holistic approach to PaaS. Rather than simply thinking about Platform-as-a-Service, we are thinking about Platform-as-a-Strategy.
With a traditional PaaS model, users request platform capabilities (servers, storage, web capabilities) on a project-by-project basis and IT provides the landscape on which users can develop their applications.
By Norm Simmonds – Consulting System Administrator, EMC IT and James Nuzzo – Senior IT Applications Development Manager
Allowing business users to create their own application landscapes with a few clicks of a button is no longer a futuristic vision. At EMC, we are developing 100 percent automated delivery of Enterprise Platform as a Service (ePaaS) using high-level cloud architecture and a self-service IT as a Service catalog.
This breakthrough in automated IT service delivery is part of EMC’s ongoing journey to the Cloud. Over the past several years, we have realized significant cost savings in transforming our IT operation into a fully automated service delivery organization.
ePaaS is the on-demand delivery of automated IT platform operations as a service, including compute, network, storage, data protection, monitoring, and application development. It allows development and IT administrators to create, modify or decommission application environments, as well as monitoring those applications via a single access point, much more rapidly than a traditional IT model.
Let us jump in feet first into ‘database as a service’. So what do we mean by this ? We have three database platforms that we can provide ‘slices’ of to our business users. Oracle and SQL Server have been the traditional platforms we have built upon and Greenplum is something we have adopted quickly and which lends itself to ‘database as a service’ very well.
How have we done this ? Tier, Consolidate, Virtualize
Of course, this has been a journey on its own merit. We started off by looking at the database tiering models required based on business criticality, required availability and I/O profiles. At EMC, we separate the mission criticalapplications and databases (as in revenue impacting and/or customer facing, typically with stringent RTO/RPO and data loss constraints) from the business critical applications and databases (impacting subprocesses vs enterprise processes).
To gain efficiencies of scale, we decided to consolidate mission-critical Oracle and business critical into 3 and 8 node Oracle grid architecture (and along the way reduced the number of Oracle versions from 9 down to 1). We also consolidated and virtualized a number of production and non-production databases for the business critical side. This consolidation and virtualization exercise resulted in the reduction of databases and database servers from 50+ to single digits. This has provided the basic technology foundation for implementing database as a service on the Oracle platform. The current environment provides us a mechanism by which a large environment can be sliced to service different needs at different points of time, with standardized and published service levels and predictable scalability and performance.
In the last post, we talked about how the ultimate nirvana is for the business to do self-service provisioning of services from a service catalog. This is somewhat easy to think about when we consider compute, storage and network services – the virtual bare metal gets provisioned within a short period of time and is made available for the user. The SLAs are typically around uptime, recovery point objective (RPO) during failure, recovery time objective (RTO) after failure, etc. This concept becomes more interesting when we talk about PaaS.
The Platform in PaaS refers to a set of capabilities that start with a base of (mostly) software components sitting above the virtual bare metal – examples we have talked about include basic capabilities such as a database, web server and application server but also extend into larger value-added capabilities such as application frameworks (Spring, .Net, force.com), content management, integration, security and information lifecycle management.
Now, once the slice of the platform is provisioned, who can configure or program to the platform ? Continue reading