xrpl-dev-portal/content/concepts/ledgers/ledger-structure.md

html, parent, blurb

html	parent	blurb
ledger-structure.html	ledgers.html	A closer look at the elements of an individual ledger block.

Ledger Structure

The XRP Ledger is a blockchain, which means it consists of a history of data blocks in sequence. A block in the XRP Ledger blockchain is called a ledger version or a ledger for short.

The consensus protocol takes a previous ledger version as a starting point, forms an agreement among validators on a set of transactions to apply next, then confirms that everyone got the same results from applying those transactions. When this happens successfully, the result is a new validated ledger version. From there, the process repeats to build the next ledger version.

Each ledger version contains state data, a transaction set, and a header containing metadata.

{{ include_svg("img/ledger.svg", "Diagram: A ledger consists of a header, transaction set, and state data.") }}

State Data

{{ include_svg("img/ledger-state-data.svg", "Diagram: A ledger's state data, in the form of various objects which are sometimes linked like a graph.") }}

The state data represents a snapshot of all accounts, balances, settings, and other information as of this ledger version. When a server connects to the network, one of the first things it does is download a full set of the current state data so that it can process new transactions and answer queries about the current state. Since every server in the network has a full copy of the state data, all data is public and every copy is equally valid.

The state data consists of individual objects called ledger entries, stored in a tree format. Each ledger entry has a unique 256-bit ID that you can use to look it up in the state tree.

Transaction Set

{{ include_svg("img/ledger-transaction-set.svg", "Diagram: A ledger's transaction set, a group of transactions placed in canonical order.") }}

Every change made to the ledger is the result of a transaction. Each ledger version contains a transaction set which is a group of transactions that have been newly applied in a specific order. If you take the previous ledger version's state data, and apply this ledger's transaction set on top of it, you get this ledger's state data as a result.

Every transaction in a ledger's transaction set has both of the following parts:

• Transaction instructions showing what its sender told the ledger to do.
• Transaction metadata showing exactly how the transaction was processed and how it affected the ledger's state data.

Ledger Header

The ledger header is a block of data that summarizes a ledger version. Like the cover of a report, it uniquely identifies the ledger version, lists its contents, and shows the time it was created, along with some other notes. The ledger header contains the following information:

• {{include_svg("img/ledger-index-icon.svg", "", classes="floating-diagram")}} The ledger index, which identifies the ledger version's position in the chain. It builds on the ledger with an index that is one lower, back to the starting point known as the genesis ledger. This forms a public history of all transactions and results.
• {{include_svg("img/ledger-hash-icon.svg", "", classes="floating-diagram")}} The ledger hash, which uniquely identifies the ledger's contents. The hash is calculated so that if any detail of the ledger version changes, the hash is completely different, which makes it also like a checksum that shows that none of the data in the ledger has been lost, modified, or corrupted.
• {{include_svg("img/ledger-parent-icon.svg", "", classes="floating-diagram")}} The parent ledger hash. A ledger version is largely defined by the difference from the parent ledger that came before it, so the header also contains the unique hash of its parent ledger.
• {{include_svg("img/ledger-timestamp-icon.svg", "", classes="floating-diagram")}} The close time, the official timestamp when this ledger's contents were finalized. This number is rounded off by a number of seconds, usually 10.
• {{include_svg("img/ledger-state-data-hash-icon.svg", "", classes="floating-diagram")}} A state data hash which acts as a checksum on this ledger's state data.
• {{include_svg("img/ledger-tx-set-hash-icon.svg", "", classes="floating-diagram")}} A transaction set hash which acts as a checksum on this ledger's transaction set data.
• {{include_svg("img/ledger-notes-icon.svg", "", classes="floating-diagram")}} A few other notes like the total amount of XRP in existence and the amount the close time was rounded by.

A ledger's transaction set and state data are unlimited in size, but the ledger header is always a fixed size. For the exact data and binary format of a ledger header, see Ledger Header.

Validation Status

{{ include_svg("img/ledger-validated-mark.svg", "Diagram: A ledger's validation status, which is added on top of the ledger and not part of the ledger itself.") }}

When a consensus of validators in a server's Unique Node List agree on the contents of a ledger version, that ledger version is marked as validated and immutable. The ledger's contents can only change by subsequent transactions making a new ledger version, continuing the chain.

When a ledger version is first created, it is not yet validated. Due to differences in when candidate transactions arrive at different servers, the network may build and propose multiple different ledger versions to be the next step in the chain. The consensus protocol decides which one of them becomes validated. (Any candidate transactions that weren't in the validated ledger version can typically be included in the next ledger version's transaction set instead.)

Ledger Index or Ledger Hash?

There are two different ways of identifying a ledger version: its ledger index and its ledger hash. These two fields both identify a ledger, but they serve different purposes. The ledger index tells you the ledger's position in the chain, and the ledger hash reflects the ledger's contents.

Ledgers from different chains can have the same ledger index but different hashes. Also, when dealing with unvalidated ledger versions, there can be multiple candidate ledgers with the same index but different contents and therefore different hashes.

Two ledgers with the same ledger hash are always completely identical.