RESTful API for orchestrating the restoration of time-based Elasticsearch snapshots.
This API gives you a simple interface for restoring snapshots based on a given time range.
For example, if snapshots are stored in the following daily (repo/snap/index) pattern: logs-%Y/logs-%Y-%m-%d/logs-v1-%Y-%m-%d then requests made to the POST /{start}/{end} route would trigger a restore of indices in the range that maps to the pattern.
High level of what the API does:
- Accepts requests to restore indices from snapshot repository.
- Incoming requests are queued for restore and queue is processed every 2 seconds.
- Requests for additional index restorations are queued while other recoveries are in progress.
- Available index snapshots are listed via the ES Snapshot/Restore API.
- Ongoing recoveries and online indices are obtained via the ES Index API.
The OpenAPI 2.0 spec is here: ./swagger.yml
ESIO Flags:
--es-host= Elasticsearch Host [$ES_HOST]
--max-restore= Maximum number of indices allowed to restore at once, default is 1 [$MAX_RESTORE]
--resolution= Resolution of indices being restored (day, month, year) [$INDEX_RESOLUTION]
--repo-pattern= Snapshot repo pattern (repo/snap/index), ex: logs-%y/logs-%y-%m-%d/logs-v1-%y-%m-%d,
[$REPO_PATTERN]
Use the make targets to build and test the esio-server.
make: validate the swagger spec, fetch dependencies, compile the source and run the server in the foreground.make test: Start Elasticsearch via Docker in background, start server in background and run all testsmake gen: Regenerate swagger framework fromswagger.ymlmake init-test-data: Creates indices with faux timeseries data and snapshot repo in Elasticsearch for testing.make start-elastic: start Elasticsearch and Kibanamake stop-elastic: stop Elasticsearch and Kibana