Hyperliquid Periodic ABCI State - Binary Data
Understand periodic_abci_states RMP files, their layout, and archival or snapshot access paths.
periodic_abci_states contains periodic binary state dumps. These files are archival inputs for state-aware consumers and can seed generated order-book snapshots. They are not a JSON event feed.
Access
| Availability | Access | Details |
|---|---|---|
| Archival | S3 Bucket | Historical access path for periodic_abci_states. There is no direct gRPC feed for ABCI state. |
| Related live data | StreamOrderbookSnapshots and GetOrderBookSnapshot | Snapshot availability and maximum response size vary by endpoint. |
File layout
Depending on the source object and compression state, the relative filename ends in one of these forms:
periodic_abci_states/<date>/<block-number>.rmp
periodic_abci_states/<date>/<block-number>.rmp.lz4Binary format
The payload is binary RMP (MessagePack) state. A .rmp.lz4 object must be decompressed before an RMP decoder reads it; the resulting bytes are still RMP, not JSON. Select a decoder that matches the node-state schema for the object you consume. Never pass the bytes to a JSON parser or infer a JSON structure from the filename.
Important fields
There is no stable top-level JSON field table for this dataset. The filename identifies the associated sequential block position, while the decoded binary fields depend on the matching RMP state schema. Preserve large integers, byte arrays, and map keys without lossy JSON conversion.
For order-book reconstruction, use a generated snapshot derived for that purpose and then apply ordered book diffs after its position. Do not treat an arbitrary binary state dump as the public snapshot response shape.
Related pages
- Generated order-book snapshots — Use the purpose-built public snapshot format for reconstruction.
- Raw book diffs — Apply ordered changes after a known state seed.
- L4 books — Consume the derived, order-level live view.