"local node not active but active cluster node not found" error

setup:
vault: 3 vaults servers (k8 pod) running in HA mode
vault version: v1.1.2
backend: consul (5 consul k8 pods)

our java app gets “local node not active but active cluster node not found” error sometimes, doing same call with curl multiple times doesn’t return any error

backend and vault k8 pods are healthy, meaning their cpu/ memory usage in grafana doesn’t show anything

what could be causing the issue?