eebc828778
The V1 table structure shows great improvements in performance, but if there's an old `kvstore` without rowid:s, these benefits are nullified: reorgs during writes and deletes remain expensive (even if the degradation is reduced somewhat). This PR creates the tables in a new file instead, and uses the old file as a read-only store - this has several interesting properties: * the old database is left completely untouched - this guarantees that downgrades work smooth (they'll only need to resync their missing portions) * starting sync after this PR means only a v1 database is created * v0 databases stick around - no migration is performed (for now) Future PR:s can introduce migration of the data from one database to another - a simply copy will take hours which is downtime we want to avoid - at that point, it might make sense to migrate straight to era files instead. |
||
---|---|---|
.. | ||
README.md | ||
attestation_pool.nim | ||
block_clearance.nim | ||
block_pools_types.nim | ||
block_quarantine.nim | ||
blockchain_dag.nim | ||
exit_pool.nim | ||
spec_cache.nim | ||
statedata_helpers.nim |
README.md
Consensus object pools
This folder holds the various consensus object pools needed for a blockchain client.
Object in those pools have passed the "gossip validation" filter according to specs:
- blocks: https://github.com/ethereum/eth2.0-specs/blob/v1.0.1/specs/phase0/p2p-interface.md#beacon_block
- aggregate attestations: https://github.com/ethereum/eth2.0-specs/blob/v1.0.1/specs/phase0/p2p-interface.md#beacon_aggregate_and_proof
- unaggregate attestation: https://github.com/ethereum/eth2.0-specs/blob/v1.0.1/specs/phase0/p2p-interface.md#beacon_attestation_subnet_id
- voluntary exits https://github.com/ethereum/eth2.0-specs/blob/v1.0.1/specs/phase0/p2p-interface.md#voluntary_exit
- Attester slashings https://github.com/ethereum/eth2.0-specs/blob/v1.0.1/specs/phase0/p2p-interface.md#attester_slashing
- Proposer slashins https://github.com/ethereum/eth2.0-specs/blob/v1.0.1/specs/phase0/p2p-interface.md#proposer_slashing
After "gossip validation" the consensus objects can be rebroadcasted as they are optimistically good, however for internal processing further verification is needed. For blocks, this means verifying state transition and all contained cryptographic signatures (instead of just the proposer signature). For other consensus objects, it is possible that gossip validation is a superset of consensus verification (TODO).
The pools presenet in this folder are:
- block_pools:
- block_quarantine: for seemingly valid blocks that are on a fork unknown to us.
- block_clearance: to verify (state_transition + cryptography) candidate blocks.
- blockchain_dag: an in-memory direct-acyclic graph of fully validated and verified blockchain candidates with the tail being the last finalized epoch. A block in the DAG MUST be in the fork choice and a block in the fork choice MUST be in the DAG (except for orphans following finalization). On finalization non-empty epoch blocks are stored in the beacon_chain_db.
- attestation_pool: Handles the attestation received from gossip and collect them for fork choice.
- exit_pool: Handle voluntary exits and forced exits (attester slashings and proposer slashings)