Overall Satisfaction with Google Compute Engine
We use Google Compute Engine for staging deployments in the web development department. We are operating off an external assessment that competitor services offer better pricing on production units, so we've agreed to keep Compute Engine in a low tier staging system only. We have yet to do our own audit from an internal efficiency perspective, and how that will impact pricing assessments.
- Compute Engine is gaining traction, and documentation is getting easier to find.
- Menus and services are structured more intelligently.
- The idea of poor Support from the Google brand prevents my technicians from picking up the phone.
- It's easier to find EC2 experts to consult and support mission-critical operations.
- Because GCE is not mission-critical for us, I would not be able to speak to ROI. Since our machines and instances are almost a match, once configured, there's really no difference.
- Bean counters will be able to speak to ROI.
Having interacted with several cloud services, GCE stands out to me as more usable than most. The naming and locating of features is a little more intuitive than most I've interacted with, and hinting is also quite helpful. Getting staff up to speed has proven to be overall less painful than others.
The Google brand itself for some reason dissuades my technicians from calling and using the service package. For some reason, whether for the brand or experience with Google products, Google is seen as functional but not supportive. We prefer to research and find certified contractors instead of using the support system. It would be interesting to assess the drivers behind this choice.
Google Compute Engine seems to complete in the speed of deployment, usability, training, and pricing. Azure's advantage is the market share of experts, due to active directory IT teams integrating with Windows on corporate networks, along with the Office suite of services.
In my opinion, most cloud applications should consider Google Compute Engine from a speed and pricing perspective. Of course, you should do an assessment based on what your application needs to do and how it needs to perform, but there's a machine for most kinds of deployments. If you have the expertise nearby, all the better.