RadiatorDB collection destroy failed

Logged when an operator's request to destroy a RadiatorDB collection fails, including failures after the requesting client disconnects. The collection field names the collection and force shows whether the destroy was forced. When the destroy stopped before changing anything, the attached error gives the cause, for example an unreachable node, a collection that is configured and in use, a storage conflict whose configured storage holds documents, or replication between the reachable nodes that did not catch up within RADIATORDB_COLLECTION_DESTROY_TIMEOUT. When individual nodes failed, failed_nodes and failures name them, destroyed_nodes lists the nodes where the collection was destroyed, and skipped_nodes lists the unreachable nodes that a forced destroy skipped. A failed commit without force leaves the nodes that committed earlier destroyed. When the destroy task stopped without a result, the attached error reports the task failure and no node fields are present; some nodes may already have destroyed the collection.

Check the attached error or the failed nodes, resolve the reported connection, lock, or replication problem, and destroy the collection again until no node stores it. A configured collection with its configured storage cannot be destroyed; remove its collection block from the RadiatorDB backend and restart or reload Radiator Server first. If nodes with the configured storage store documents, remove and reprovision the nodes with the different storage instead, as described in Destroy a collection. If a node is unreachable, forced destroy destroys the collection only on the reachable nodes.