Un cluster Kubernetes sur deux sites : une simple histoire de disponibilité, non ?

Slides disponibles sur Github Pages

Generate

./setup.sh

git submodule update --init --recursive

docker compose up
python3 -m http.server
# open localhost:8000

Sources

• https://etcd.io/docs/v3.5/op-guide/failures/#network-partition

• https://kubernetes.io/docs/concepts/scheduling-eviction/assign-pod-node/#affinity-and-anti-affinity

• https://docs.rke2.io/datastore/backup_restore

• https://etcd.io/docs/v3.5/op-guide/recovery/

• https://kafka.apache.org/documentation/

• https://strimzi.io/

• https://strimzi.io/docs/operators/latest/configuring.html#type-Rack-reference

• https://learn.conduktor.io/kafka/kafka-topic-configuration-min-insync-replicas/#Overriding-Topic-Configuration-Defaults-2

• https://cloudnative-pg.io/