Skip to content

Latest commit

 

History

History
144 lines (126 loc) · 9.01 KB

storage.md

File metadata and controls

144 lines (126 loc) · 9.01 KB

Storage components

As of now, RocksDB is used as a key-value database for all Storages.

1. Ledger

  • The ledger is an ordered log of transactions with each transaction assigned a unique monotonically increasing positive integer called sequence number.
  • Sequence numbers start from 1 and then increase by 1 for each new transaction. There are no gaps in sequence numbers.
  • RocksDB is used as a key-value storage where key is the sequence number and value is the transaction.
  • A transaction is serialised (currently as MsgPack) before storing in the ledger, more on this in the Serialisation doc.
  • Exact format of each transaction can be found in Indy Node Transactions.
  • Each node hosts several ledgers each identified by a unique id:
    • Audit Ledger (id is 3): Contains transactions for every ordered 3PC Batch with information about the pool state at the moment of batch ordering. It's used for synchronization between other ledgers and recovering of pool state by new or restarted nodes. It can also be used for external audit and validation of all transactions written to the ledger. See Audit Ledger for more details.
    • Pool Ledger (id is 0): Contains transactions related to pool membership, like joining of new nodes, suspension of existing nodes, changing the IP/port or keys of existing nodes.
    • Domain Ledger (id is 1): Contains transactions related to the core application logic. Currently it contains NYM transactions. The indy-node codebase extends this ledger with other identity transactions.
    • Config Ledger (id is 2): Contains transactions related to the configuration parameters for which the pool needs to agree, eg. if each node of the pool needs to use the value 5 for a config variable x, then this ledger should contain transaction specifying the value of x as 5. The indy-node codebase extends this ledger with source code update transactions.
    • More ledgers can be added by plugins.
  • Each correct node should have exactly the same transactions for each ledger id.
  • A ledger is associated with a compact merkle tree.
    • Each new transaction is added to the ledger (log) and is also hashed (sha256) and this hash becomes a new leaf of the merkle tree which also results in a new merkle root. Thus for each transaction a merkle proof of presence, called inclusion proof or audit_path can be created by using the root hash a few (O(lgn), n being the total number of leaves/transactions in the tree/ledger) intermediate hashes.
    • Hashes of all the leaves and intermediate nodes of the tree are stored in a HashStore, enabling the creating of inclusion proofs and subset proofs. A subset proof proves that a particular merkle tree is a subset of another (usually larger) merkle tree; more on this in the Catchup doc.\
    • The HashStore has 2 separate storages for storing leaves and intermediate nodes, each leaf or node of the tree is 32 bytes. Each of these storages can be a binary file or a key value store. The leaf or node hashes are queried by their number.
  • When a client write request completes or it requests a transaction with a particular sequence number from a ledger, the transaction is returned with its inclusion proof.
  • States and Caches can be deterministically re-created from the Transaction Log.
  • There are 9 storages associated with the Ledgers (3 for each of the ledgers):
    • Audit Ledger:
      • audit_transactions (Transaction Log)
      • audit_merkleLeaves (Hash Store for leaves)
      • audit_merkleNodes (Hash Store for nodes)
    • Pool Ledger:
      • pool_transactions (Transaction Log)
      • pool_merkleLeaves (Hash Store for leaves)
      • pool_merkleNodes (Hash Store for nodes)
    • Domain Ledger:
      • domain_transactions (Transaction Log)
      • domain_merkleLeaves (Hash Store for leaves)
      • domain_merkleNodes (Hash Store for nodes)
    • Config Ledger:
      • config_transactions (Transaction Log)
      • config_merkleLeaves (Hash Store for leaves)
      • config_merkleNodes (Hash Store for nodes)

Relevant code:

  • Ledger: plenum/common/ledger.py and ledger/ledger.py
  • Compact Merkle Tree: ledger/compact_merkle_tree.py
  • HashStore: ledger/hash_stores/hash_store.py

2. State

  • Each Ledger (except Audit Ledger) has a State. State is a resulting view (projection) of the ledger data used by Node and Application business logic, as well as for read requests.
  • The underlying data structure of state is the Merkle Patricia Trie used by Ethereum.
  • The state can be considered a collection of key value pairs but this collection has some properties of merkle tree like a root hash and inclusion proof of the keys, eg. If there are 3 key value pairs k1-v1, k2->v2 and k3->v3, then state will have a root hash say R and a proof can be generated that state with root hash R has a key k1 with value v1, similarly for other keys. When a new key is added or value of an old key is changed the root hash changes from R to R'. Note that the order of insertion of keys does not matter, any order of the keys will result in the same root hash and same inclusion proof.
  • It's possible to get the current value (state) for a key, as well as a value from the past (defined by a state root hash).
  • The state is built from a ledger, hence each ledger will usually have a corresponding state. State can be reconstructed from the Ledger.
  • There are 3 storages associated with every Ledger except Audit one:
    • pool_state
    • domain_state
    • config_state

Relevant code:

  • State: state/pruning_state.py
  • Merkle Patricia Trie: state/trie/pruning_trie.py

3. Node Status Database

  • Auxiliary storage to persist data needed for consensus algorithm. As of now the following information is stored there:
    • Last sent pre-prepare seqNo for primaries on backup instances to be able to recover it and continue ordering on a backup instance if the coirresponding primary is restarted.
    • Instance Change messages from every Node a given Node hasn't yet started a View Change.
  • Storage name: node_status_db

4. BLS Multi-Signature Database

  • Every transaction for the given time (that is every update of the state) is signed (using BLS schema) by all nodes during consensus, and BLS multi-signature is created and stored in this state for every state root hash.
  • This is used together with a State Proof as part of replies to read requests.
  • Storage name: state_signature

Relevant code:

  • BlsStore: plenum/bls/bls_store.py

5. Request to Transaction Mapping Database

  • This database stores the following mappings in a key value store:
    • request_payload_digest -> ledger_id<delimiter>seq_no
    • request_full_digest -> request_payload_digest
  • Each client request is uniquely identified by a payload_digest (without sigantures and plugin fields).
  • full_digest is calculated against payload, as well as multi-signatures and plugin fields.
  • When this request is ordered (consensus successfully completes), it is stored in the ledger and assigned a sequence number.
  • One use case is that the client can ask any node to give it the transaction corresponding to a request key (payload_digest), and the node will use this database to get the sequence number and then use the sequence number to get the transaction from the ledger.
  • The storage also protects against the following malicious actions:
    • writing duplicate transactions (with the same payload)
    • writing transactions with the same payload, but different multi-signatures
  • Storage name: seq_no_db

Relevant code:

  • ReqIdrToTxn: plenum/persistence/req_id_to_txn.py

6. Timestamp storage

  • This database stores the mapping timestamp -> state root hash in a key value store.
  • This is applied to transactions in Domain ledger only.
  • This is used to get the state root hash for the given timestamp to efficiently get data from the past.
  • Storage name: state_ts_db

Relevant code:

  • StateTsDbStorage: storage/state_ts_store.py

7. Identifier Cache Database (in Indy Node)

  • This database stores the mapping DID -> verkey/role in a key value store.
  • It's used for efficient validation and getting verkey for a DID.
  • Storage name: idr_cache_db

Relevant code:

  • IdrCache: indy-node/persistence/idr_cache.py

8. Attribute Database (in Indy Node)

  • Raw and encoded attributes from ATTRIB transaction are stored in this database.
  • The ATTRIB transaction in the Domain Ledger contains the hash of the data only.
  • Storage name: attr_db

Relevant code:

  • AttributeStore: indy-node/persistence/attribuite_store.py

Common Storage abstractions

The data structures in the code use some abstractions like a KeyValueStorage interface which has a LevelDB and RocksDB implementations, as well as file implementations (both single and chunked file storages).

Relevant code:

  • storage/kv_store.py
  • storage package