“When Smartson decided to improve the site speed by introducing a cache server we also took the opportunity to try out Compute Engine which is a part of Google Cloud Platform. In less than 30 minutes we had dimensioned the server, created the server instance, installed the operating system and the cache software, Varnish, including configuring the cache software.“. Says Fredrik Westermarck, CTO of Smartson.
Requirements and background
Smartson is, despite it's 14 years in the industry a relatively small company. Withlimited resources in terms of time and with an ambition to continuouslyintroduce improvements in services, such as web and the tools that we use fromthird parties requires that the solutions are simple to manage and maintain.
Amazon Web Services
We have previously used Amazon Web Services in other solutions.AWS is obviously the first choice when you need access to storage and virtual servers in the cloud.However we felt that the administation interface wasn't that intuitive in some repects and we realized that in order for us to have easy maintainable solution we would need to invest some time on the configuration of AWS. For instance by creating scripts to manage both AWS S3 and AWS EC2. At the time we didn't have the time required to develop a method that suited us.
Google Cloud Platform
The next solution that we chose to evaluate was Google Cloud Platform which is a direct competitor to Amazon Web Services. The reason, for us, that Google Cloud Platform was our second choice is because we are already using Google Apps for Work to manage our users and to store documents and e-mail. To get up and running all we had to provide our billing information. Initially Google grants you 300 US$ to spend over a couple of months. The offer is useful and should be enough to get you up and running to try out your use-case.
When the billing information is completed, you can immediately start towork against Googles APIs, Compute Engineor, as in our case, Container Engine.
To create a server instance, configure the available hardware, add the instances storage capacityand to install an image went quickly and smoothly. Much faster than AWS where itis a plethora of images to choose from. The virtual server instance started and was ready to get it's software installed.
The software was installed and configured in the usual manner. It took lessthan 20 minutes for us to have a virtual server instance that was ready tobe inserted into production, I think that's pretty good.
Google Compute Engine provides excellent tools for monitoring and managing alarms of thevirtual hardware and the software in the setup with Google Cloud Monitoring. All in one simple interface. There's also an app to monitor and manage instances from your phone or tablet, it is quite convenient to have the ability to restart an instance or apply security updates, from the sofa or bed in the evening. I can attest to that myself.
It is worth to try out Google Cloud Platform as an alternative to AWS ifyou're not already using AWS and are about to launch a virtual server orserver environment into production.