Realtime reports and Historical reports are inaccessible after node outage

Last Updated : Nov 01, 2021 |
Prolog information

Condition

Users are unable to access Realtime or Historical reports, and any URL using the Cluster FQDN is inaccessible.

Cause

IPVS routing table are stalled due to a down node.
No HTTP request are routed from clusterIP to ingressgateway endpoints.

Solution

  1. For Avaya Analytics™ release 4.0.0.1 and 4.1.1.0, perform following steps:
  2. Login to CCM as root
  3. Get the names of the kube-keepalived-vip pods
    ADDITIONAL INFORMATION: kubectl get pods | grep common-services-kube-keepalived-vip
  4. Restart each of the kube-keepalived-vip pods
    ADDITIONAL INFORMATION: kubectl delete <pod_1> <pod_2> <pod_3>
    STEP RESULT:
    Once the kube-keepalived-pods is restarted and running, this issue is resolved. Realtime Reports and Historical Reports should be accessible.