I had deployed eck 1.3.0 in GKE cluster. I saw pdb was refreshing even though there were no changes in any elasticsearch cluster nor pdb related configurations. Can anyone please tell why would this happen.
The operator ensures the PDB allows one Pod to be disrupted is Elasticsearch reports a green health, zero otherwise (yellow, red, or unknown health).
It continuously polls the Elasticsearch health and recreates the PDB accordingly.
In your case I guess the PDB recreation matches a change in Elasticsearch health.
See also the official doc: Pod disruption budget | Elastic Cloud on Kubernetes [1.6] | Elastic.
Thanks for the reply. Got it.