Last week was Serverless Conf and I have to admit I completely understand why in just one short year this event has gone from 100 attendees to this year’s 450. It’s clear that developers are getting the memo about this burgeoning technology whose moniker (Serverless) may not actually be founded in 100% truth, but to a developer, the message is received.
A developer can create mobile, web and/or IoT applications without the looming decisions they used to face such as:
Instead, somewhere, in a data center in “never never land” resides that ‘server’ whose provisioning needs are taken off the developer’s shoulders…..
And just like that, the developer has become server – less.
It was fascinating to talk to so many different people from different industries to find out what they’re trying to do with the serverless framework because it is clear that the possibilities are endless and the dreams are big. I was also able to share the news that IBM has their own serverless platform, IBM Bluemix OpenWhisk which runs on the open source project, Apache OpenWhisk. This was a great opportunity for IBM to gain traction within the rapidly growing market.
So what else is interesting about serverless computing and how does OpenWhisk shine?
I would like to highlight a few examples of how I think serverless (specifically Openwhisk) is changing the app development game:
IoT + OpenWhisk = Open your wallet, it’s about to get full
In a world of connected devices, managing that traffic flow and being able to make sense out of all the data coming in is critical for a business to make decisions rapidly and adjust to meet the demands of their clients or optimize their operations. A OpenWhisk user, GreenQ, is a self proclaimed trailblazer in the area of IoG, or Internet of Garbage. Yep, its a thing – or is now anyway. GreenQ is a waste collection organization that installed sensors on their trucks and waste bins to ensure the most optimal route for collection as well as real time analysis on waste bin weight and time before collection. By integrating the sensor data with OpenWhisk, they were able to process the messages coming in real time and make adjustments immediately. Ultimately this reduced their operating costs by nearly 50% when they tapped into OpenWhisk’s autoscale capabilities, plus its ability to take action on an event (each lift of a waste bin), which in turn, kicks off an action (capture bin data and stores in a database). This combination of autoscaling and event triggers gave GreenQ the power it needed to provide its clients with intelligent, highly efficient waste collection at a fraction of the cost.
Repeatable and redundant tasks + OpenWhisk = Don’t blink, you’ll miss it
What if a teller did not have to get involved in check deposits? Well they don’t, thanks to technology like OpenWhisk. OpenWhisk is able to compile information provided from something like OCR, trigger an event and automatically deposit that check in a millisecond. Something that may have required minutes previously for a human to complete each check, now is done in a blink. OpenWhisk just took a repeatable action and cut it down to a small fraction of time required to complete that task. I can only imagine the ways the enterprise would want to deploy this kind of time/cost savings in any way possible!
- Cognitive services + OpenWhisk = Take a nap, the work is done for you
Cognitive, or machine learning, is gaining some serious steam for obvious reasons. These type of services have broadened the scope of what we knew was ever possible in the areas of a machine being able to critically think through a problem and come to logical conclusions. By combining this revolutionary technology and the power of Openwhisk, what was once impossible, now becomes just impossibly fast. Take for example, computers take in millions of images every day and IBM Watson Visual Recognition is able to take these images and be taught not only what is in the image but the context around the graphic. By leveraging OpenWhisk, a new image is treated as an event that calls on this Watson service to identify and catalog. What was once a pile of data without any meaning is immediately transported to valuable information that can be aggregated to make decisions in real time. With OpenWhisk in the driver’s seat, telling an IBM Watson service to do its thing, the user can sit back and relax and let the analytics unfold before them.
So to wrap this up, basically serverless is changing the game in application development, just like cars changed the game for movement.
People walked or rode horses eveywhere before cars, they had to worry about things like the proper water/food for the trip. They had to prepare for some heavy rain or dry heat. Then came the car. Having the ability to push a pedal to get yourself from point A to point B eliminated the need to spend time thinking about what ‘could’ happen and now the driver could think only of where did they want to go and what sights were around the corner. Serverless has done this same thing for the developer. No longer does the developer need to worry about server capacity, unpredictable traffic flow or even who might access that server. The developer now gets to spend time thinking about where they want to go and what innovation is just around the corner.
So no, serverless does not mean no servers, it simply means the developer gets to develop, server – less.