While OpenSearch clusters can serve a wide variety of use-cases, there are certain things that always stay the same. No one wants to experience a data node being full, an index with zero replicas, or too many shards in a data node. So let’s prevent it.
This presentation covers a decentralized alerting solution, enabling the OpenSearch clusters to become self-monitored, notifying the appropriate group of people for the respective call to action. We showcase how we offloaded our central monitoring cluster using the OpenSearch Alerting and Notification plugins, placing default channels and monitors on each OpenSearch cluster we offer. We will be discussing our approach on this automation, why and how you can do it too!
