98 points by k8s_geek 6 months ago flag hide 15 comments
theawsrefugee 6 months ago next
After battling with AWS for a while and hitting some limitations with scaling and cost management, I decided to migrate my startup to Kubernetes.
awsfreeuser 6 months ago next
What specific AWS issues did you experience? I'm thinking of migrating my personal project off AWS due to similar reasons.
satisfiedk8ser 6 months ago next
The biggest issue was the difficulty of managing AWS resources at scale. Our DevOps costs skyrocketed as we added more microservices and grew our user base. CI/CD pipelines also became a challenge to manage and Kubernetes simplifies this drastically with kaniko and config management tools.
cloudninja35 6 months ago prev next
Kubernetes really shines in terms of resource efficiency and scalability. It's also platform agnostic which means you can use any cloud provider or even on-prem infrastructure.
bettertogether 6 months ago next
It's a common misconception that K8s is hard to manage. Sure, it can be overwhelming at first, but it gets much easier once you're familiar with the pods, deployments, statefulsets, ingresses, and other primitive concepts. K8s handles rolling updates and rollbacks seamlessly, and you can update and scale apps without downtime. I think it's definitely worth the investment.
newsolutions 6 months ago next
We integrated Jenkins X into our stack to streamline CI/CD management with K8s and it's been great ever since! We also found Rancher Kubernetes Management to be very helpful for our team.
cloudninja35 6 months ago next
Agreed, Jenkins X is really powerful but make sure to allocate time to understand and optimize its config settings, especially for updating plugins and pipelines. It took some time for us to fine-tune it to our needs.
devopsguru 6 months ago prev next
I strongly agree with betterTogether. K8s gets much better once you get past the initial learning curve, and its benefits in infrastructure abstraction and resource orchestration pay off. By the way, what did you use for monitoring and logging with K8s?
satisfiedk8ser 6 months ago next
For monitoring, we use Prometheus and Grafana, while for logging we use Loki and Grafana. They've simplified observability in our devOps duties.
awsfreeuser 6 months ago next
Thanks, satisfiedK8ser, that's informative. I tried using Prometheus but wasn't impressed due to the short retention period. Will look into Loki and Grafana.