Item: Google Kubernetes Engine
Rating: 8
Author: Verified User

Overall Satisfaction with Google Kubernetes Engine

Use Cases and Deployment Scope

We have an application for creating Internal reports from Billing Data for customers. We have deployed that application to Google Kubernetes Engine through GitLab. Also, we provide an automatic secured Kubernetes offering to our customers in Germany.

Pros and Cons

Pros

Autoscaling Application to mitigate the increase in traffic.
Automatically roll out upgrades to applications.
As it is available through all major Cloud providers, your applications can run anywhere.
Configuration management. Automate the deployment of the application.

Cons

Sometimes, it's not the logical offering if your application is not complex. Managed offerings like App Engine are better in some cases. So, it could have a lightweight offering with free tiers.
For Machine Learning, there are free resources like labs to test. Doing that on Kubernetes is a very tedious task.
It has already improved to be a managed service. But still, developers cannot deploy on Kubernetes without hassle.

Return on Investment

It has given us flexibility and reduced deployment timelines by half.
With the help of Git, it's really good that we can maintain the state of our application. And update or rollback features, which was very difficult earlier.
Our application is for internal use right now, but we have used GKE as we will open it to the world. And their autoscaling will be a necessity.

Usability

It's a great product if you learn it. It has flexibility and is very strong. Autoscaling and Resource management make running huge applications a breeze. Using Helm with Kubernetes and Terraform for infrastructure creation can totally automate your CICD pipeline. You also get easy access to CUDA cores for machine learning.

Alternatives Considered

Helm, HashiCorp Terraform and GitLab

We have a CICD pipeline, which we wrote using the GitLab CI file. This is connected directly to our GKE cluster. So, any change in our code will directly start the CICD pipeline. The pipeline first tests the deployment on testing environments. We are also using Helm charts to make Kubernetes Deployments reusable.

Key Insights

Do you think Google Kubernetes Engine delivers good value for the price?

Yes

Are you happy with Google Kubernetes Engine's feature set?

Yes

Did Google Kubernetes Engine live up to sales and marketing promises?

Yes

Did implementation of Google Kubernetes Engine go as expected?

Yes

Would you buy Google Kubernetes Engine again?

Yes

Other Software Used

Helm, HashiCorp Terraform, Prometheus, GitLab

Likelihood to Recommend

If your application is complex, if it's planet-scale, or if you need autoscaling, then Kubernetes is best suited. If your application is straightforward, you can opt for App Engine or Cloud Run. In many cases, you can prefer to run the cloud on GKE. But once you deploy on Kubernetes, you get the flexibility to try different things. But if you don't seek flexibility, it's not an option for you.

Google Kubernetes Engine Feature Ratings

Evaluating Google Kubernetes Engine and Competitors

Products Replaced

Yes - App Engine, Cloud Run, VM

Key Differentiators

Scalability
Integration with Other Systems
Ease of Use

We needed a robust solution for our Application which can scale easily. We had already upgraded App Engine image for more compute but still our application needed more as we use multi-threading, pandas etc.
Kubernetes was a best option here. It also easily integrated with our existing products. So it was easy to chagne to GKE.

Evaluation Lessons Learned

First we will try deploying on Kubernetes Autopilot, then GKE as we already have deployed on Standard GKE we cannot try autopilot for now. We could have better judged what other options we have. As we are a Google reseller, we could only use Googles offering initially. But we could have checked Amazon and Azure offerings also.

Comments

Please log in to join the conversation

GKE: Powering Kubernetes Workloads at Scale.