As I write this post from DreamForce, my second year attending the event, the buzz here is incredible. The last two days have seen a slew of announcements from Salesforce.com. Being a showman is definitely not something Marc Benioff would complain being tagged as.
Force.com for all the accolades still carried the tag of being proprietary, built around a proprietary technology Apex and proprietary database. Salesforce.com shed part of stigma with the announcement of VMForce last year, opening up the platform for Java Developers albeit just the Spring faction. Today they took another big step when they announced the acquisition of Heroku, the leading Ruby on Rails Platform. With startups increasingly using Ruby on Rails as their platform of choice to build out social, mobile and web2.0 applications, I think this is fantastic acquisition for Salesforce.com.
Heroku is a multi-tenant platform for deploying Ruby/Rails applications. It runs on top of Amazon’s EC2, while abstracting all those details from users. Heroku offers both shared and dedicated deployment options and very flexible usage-based pricing. Looking at their site they claim running 100K applications on their platform with amongst others companies like Best Buy. In addition, they have their own version of marketplace for third party developers to publish and sell applications built on Heroku. You can find solutions for MongoDB, Memcache, Redis and many Facebook, Twitter applications.
Heroku is as geeky as geeky gets. And I am not talking about their technology. Their metering unit is called Dyno. A dyno is a single web process running your code and responding to HTTP requests. You buy more dynos as you need more scale. In a traditional Infrastructure-as-a-service, you would provision more VMs to crank up the capacity, in Heroku you increase the number of dynos spread across a grid. A single dyno can handle 10-100 requests per second. Heroku also comes with other essential ingredients of a scalable platform like HTTP caching using Varnish, Background/Job Queues, Image Caching. Other technologies available in the platform include , Full Text Search (Solr), a full-featured PostgreSQL DBMS and NOSQL database choices through its marketplace.
A quick glimpse at the Heroku offering details
Salesforce.com was one of the first companies that saw that there was a need for a cloud platform, even as they continued to charge forward with their Sales Automation SaaS application. While it took a while for companies to become believers of the virtues of Platform-as-a-Service, now it is absolutely clear that it will be the largest cloud service amongst all the XaaS segments.
Take this acquisition announcement alongwith the announcement of the announcement of Database Service(which I will cover in a separate post), the RemedyForce offering for IT Service Management in partnership with BMC and the past VMForce announcement and you now get a clear picture of the full platform around Force.com evolving.
So what are the pluses?
- Adding the Ruby on Rails capability to the application development platform is definitely a big plus.
- Creating the database as a service is definitely a step in the right direction. While the database might not match up well with the existing on-premise databases today, making it a separate offering will ensure the roadmap will be evolve and the database can only get better from here on.The learnings from Heroku team in terms of supporting NOSQL databases which might be net addition to database.com.
- Salesforce.com has primarily been focused on enterprise. Adding Ruby on Rails gives it access to startups and other edge application domains like social applications.
Does this mean Salesforce.com is ready world dominance? Far from it. Here are some reasons why
- Ruby on Rails is not the predominant development environment in the mainstream IT yet. Most of the companies have at best built some edge applications like social applications. It is still making its way out of the geekdom to the mainstream. The tooling around Ruby on Rails is still in the nascent stage. Having a successful company like Salesforce.com put its muscle behind it can only make it better.
- The focus for both companies have been edge applications. For Salesforce.com to make real dent into corporate IT and initiate any kind of exodus will need to outline how companies can reuse existing business logic, process and easily move existing applications to Cloud.
- While the effort to become more open is great, a platform is truly open when it allows companies to not just move their applications from corporate datacenter to their platform but also to take it to another platform. In a way they should demonstrate there is no lock-in. VMForce and Heroku should alleviate that concern, but the journey has just started.
- Heroku is entirely based on Amazon EC2. Force.com on the other hand has its own infrastructure. It would be interesting to see what will be done there. Force.com has ways to go before they can match the tooling, management that Amazon has. Heroku is already considered expensive and the pricing might get more expensive if Heroku transitions to Force.com infrastructure.
VMForce started off with a bang and since then we have not heard a great lot about the successes. The more successes they can demonstrate the more they will get considered as true platform. They have gotten off to a great start.