2019-06-26 Kubernetes Subgroup Meeting notes

Wayne Schneider , Jakub Skoczen

Meeting notes

Kubernetes discussion at the FOLIO face-to-face:

• Project has settled on Kubernetes as reference production installation
• Development issues that will get priority:  OKAPI-729 - Getting issue details... STATUS ,  OKAPI-736 - Getting issue details... STATUS  – see umbrella issue  UXPROD-1823 - Getting issue details... STATUS

Update on  FOLIO-1550 - Getting issue details... STATUS

John Malconian

Meeting notes

John Malconian has set up some clusters in the FOLIO AWS environment, including HA Rancher and an initial cluster config. He has created some automation around preparing VPCs for cluster installation.

Chicken and egg issue – how to set up a cluster for HA Rancher without Rancher – used Rancher toolkit to build a cluster on EC2 instances.

Cluster for development has been set up, along with RDS for dev.

Cluster nodes are m4.xlarge (4-core) – this is open for experimentation. Expectation is that the CI environment will need to run a lot of nodes in order to run a lot of modules. Memory may be the big constraint.

How to monitor FOLIO modules themselves? What are options for integration with built-in Rancher Prometheus monitoring? How can we leverage the Prometheus and JMX features that are built into the standard container image? What about Carbon metrics and Grafana visualization that is already instrumented (in Okapi)?

Out of the box, enabling Prometheus/Grafana on a workload in Rancher already gathers many useful metrics for performance and load testing.

Wayne Schneider

Meeting notes

For next week: network topology on AWS (John Malconian), with pictures

Review documentation created as part of  FOLIO-2054 - Getting issue details... STATUS