0.6.1

  • Adds a KV partition listing for slap_inspect that reads keys only and includes the shard's ownership-confirmation row. Malformed keys appear with an :unknown_key diagnostic.

0.6.0

  • Uses slap_slatedb 0.4.0, which upgrades SlateDB from 0.17 to 0.18. Rollback to SlateDB 0.17 is not guaranteed; see the upgrade precautions.

0.5.0

  • Each shard now limits the requests and bytes waiting for durability from writes and linearizable reads. A request that exceeds current capacity returns {:error, :overloaded} (HTTP 503). One that exceeds a byte limit by itself returns {:error, {:bad_request, :request_too_large}} (HTTP 413). Rejected requests are not applied. Configure the limits in the cluster's child_options; see Limits on pending work.
  • Adds Slap.KV.batch/3 and JSON POST /v1/kv/{partition} for atomic puts, deletes and condition checks within one partition. Batches default to at most 1,000 operations and 8 MiB of encoded keys and values. Larger batches return {:error, {:bad_request, :batch_too_large}} (HTTP 413).
  • Adds read-only partition inspection for Slap Inspect.
  • Adds telemetry events for rejected requests and pending work; see Slap.KV.Telemetry.

0.4.0

  • Slap.KV.scan/2 accepts cursor: nil to start at the beginning of the selected range. A returned cursor: nil marks the end of iteration. prefix: nil, gte: nil, and lt: nil remain invalid.

0.3.0

  • Slap.KV.Cluster sets the default SlateDB flush_interval to 10 ms. Pass settings: %{flush_interval: "100ms"} to restore the previous default.
  • With probe: :background, the cluster starts before its store is available. Operations return {:error, :unavailable} until its shards open.

0.2.0

  • Uses slap_slatedb 0.2.0, which upgrades SlateDB from 0.16 to 0.17. Databases written by 0.1.0 open unchanged.

0.1.0

Initial release of slap_kv.