78 points by pghyperscale 6 months ago flag hide 6 comments
coder1 6 months ago next
Fascinating read about hyperscaling PostgreSQL with Kubernetes in production! I've been curious about the benchmarks and how this setup handles read-write workloads.
dbengineer2 6 months ago next
(reply to coder1) The paper provides fantastic details on benchmarks. I recommend checking out the best practices for managing read-write workloads in their setup. It helps a long way in achieving consistent performance.
maxscalability 6 months ago prev next
(reply to coder1) For read-write workloads, the key is to horizontally scale the resources in a manner so that your PostgreSQL Pods make the most of the available Kubernetes resources. This way, you prevent any under-utilization or resource starvation.
k8spro 6 months ago prev next
Great share! I work with Kubernetes daily, and the post highlights some impressive architectures and optimizations for PostgreSQL.
infraarchitect 6 months ago next
(reply to k8spro) Absolutely! I was particularly impressed with how they managed stateful sets for data durability ensuring that PostgreSQL workloads were stable and resilient in production.
k8sadmin3 6 months ago next
(reply to infraarchitect) Yes, stateful sets enable a more seamless management of PostgreSQL instances. However, you must be vigilant about cache management and data distribution to avoid potential bottlenecks.