Trying to setup Curator to run automatic snapshots of old indices in my cluster, but I cannot get it to recognize an existing repository to start the snapshot process.
On my test cluster, I have successfully created a repository called "Archive"
GET _cat/repositories?v returns:
id type
Archive fs
I've even successfully created a snapshot:
GET _snapshot/Archive/snapshot_1 returns:
{ "snapshots": [ { "snapshot": "snapshot_1", "uuid": "p6TN8K2qRGaBlTzoRSfv6A", "version_id": 5040199, "version": "5.4.1", "indices": [ "winlogbeat-2017.05.11", "winlogbeat-2017.05.12" ], "state": "SUCCESS", "start_time": "2017-11-30T19:51:31.691Z", "start_time_in_millis": 1512071491691, "end_time": "2017-11-30T19:52:18.587Z", "end_time_in_millis": 1512071538587, "duration_in_millis": 46896, "failures": [], "shards": { "total": 10, "failed": 0, "successful": 10 } } ] }
But if I try to run a snapshot from Curator, I get the following in my logs:
2017-11-30 15:29:21,996 INFO Preparing Action ID: 1, "snapshot" 2017-11-30 15:29:22,007 INFO Trying Action ID: 1, "snapshot": snapshot old indices to archival storage 2017-11-30 15:29:26,256 ERROR Failed to complete action: snapshot. <class 'curator.exceptions.ActionError'>: Cannot snapshot indices to missing repository: Archive
Action file looks like this:
actions: 1: action: snapshot description: snapshot old indices to archival storage options: repository: Archive wait_for_completion: True
What am I missing? Is there a way I need to call out the repository name to get Curator to recognize it?