The replication.mode sync option instructs to create one or more cluster members as synchronous replicas. As indicated in the CRD reference the cluster will not block transactions in the leader in the event of replica lost, since Patroni will turn off the synchronous replication.
Lets move forward to the next step by creating a synchronous replication cluster:
$ cat << EOF | kubectl apply -f -
apiVersion: stackgres.io/v1
kind: SGCluster
metadata:
namespace: failover
name: sync-cluster
spec:
postgres:
version: '16.1'
instances: 3
sgInstanceProfile: 'size-s'
pods:
persistentVolume:
size: '10Gi'
configurations:
sgPostgresConfig: 'pgconfig1'
sgPoolingConfig: 'poolconfig1'
prometheusAutobind: true
nonProductionOptions:
disableClusterPodAntiAffinity: true
replication:
mode: sync
role: ha-read
syncInstances: 1
EOF
Since syncInstances is set to 1, one synchronous replica and one asynchronous replica are created accordingly.
$ kubectl -n failover exec -it sync-cluster-0 -c patroni -- bash - patronictl list
+ Cluster: sync-cluster (7369946595341132525) -----+-----------+----+-----------+
| Member | Host | Role | State | TL | Lag in MB |
+----------------+------------------+--------------+-----------+----+-----------+
| sync-cluster-0 | 10.244.0.21:7433 | Leader | running | 1 | |
| sync-cluster-1 | 10.244.0.23:7433 | Sync Standby | streaming | 1 | 0 |
| sync-cluster-2 | 10.244.0.25:7433 | Replica | streaming | 1 | 0 |
+----------------+------------------+--------------+-----------+----+-----------+