Protecting critical information of an application orchestrator deployment, such as an etcd data store in a kubernetes cluster

A computer-implemented method comprising:

• by a storage manager executing on a first computing device comprising one or more hardware processors and operating outside of a production cluster that comprises an application orchestrator computing environment, the production cluster comprising an etcd data store that stores configuration data describing objects deployed in the production cluster:
• initiating a backup operation of the etcd data store according to a first data protection plan associated with the etcd data store;
• by a coordinator data agent executing outside of the production cluster and communicatively coupled to the production cluster, instructing the production cluster to generate an etcd snapshot of the etcd data store and to store the etcd snapshot at a designated data path within the production cluster;
• responsive to determining that the etcd snapshot is not accessible at the designated data path, causing a temporary data transfer agent to be deployed as a containerized application in a worker pod executing within the production cluster, wherein the temporary data transfer agent is configured to create the designated data path within the production cluster;
• by the temporary data transfer agent, accessing the etcd snapshot at the designated data path and transmitting contents of the etcd snapshot to the coordinator data agent;
• by the coordinator data agent in conjunction with a media agent operating outside of the production cluster, generating a secondary copy of the etcd data store based on the contents of the etcd snapshot and storing the secondary copy at a secondary storage that is outside of the production cluster;
• and
• based on successful storage of the secondary copy at the secondary storage, causing the worker pod hosting the temporary data transfer agent to be torn down within the production cluster.