CHANGELOG.md

0.8.1:

• do not repeat summarization for epochs

0.8.0:

• support deneb
• add t_blob_sidecars for Deneb
• add f_blob_kzg_commitments to t_blocks for Deneb
• add f_blob_gas_used to t_block_execution_payloads for Deneb
• add f_excess_blob_gas to t_block_execution_payloads for Deneb

0.7.7:

• add attestations provider
• do not attempt to summarize ahead of finalized epoch
• update PGX to version 5

0.7.6:

• Fix error in the Blocks() provider

0.7.5:

• Add provider for epoch summaries
• Add withdrawals to epoch summaries table
• Add withdrawals to validator day summaries table
• Add excess data gas to block execution payload table

0.7.4:

• Ensure that only canonical withdrawals are included in summaries

0.7.3:

• API update only

0.7.2:

• fix upgrade issue from significantly earlier version of database

0.7.1:

• fix issue where capella fork was not detected
• pick up capella epochs in finalizer
• add validator index to BLS to execution changes (thanks to @samlaf)
• add withdrawal credentials to validators (thanks to @samlaf)

0.7.0:

• speed up sync by only updating changed validators
• add open telemetry tracing
• avoid fetching duplicate proposer duties
• introduce daily validator summaries, and pruning of validator balances and validator epoch summaries
• support Capella

0.6.17:

• expose read-only transaction functions

0.6.16:

• tidy up summarizer error messages on failures
• use read-only transactions where applicable

0.6.15:

• catch edge case where summarizer may not include all attestations first time around
• warn if running against a newer version of the database schema

0.6.14:

• avoid crash if no response is given to a 'eth_getBlockByHash' call
• do not refetch beacon committees if we already have them

0.6.12:

• periodically update the information obtained by the spec service

0.6.11:

• add configurable database connection pool size
• separate read-only and read-write transactions internally
• share activity semaphore between blocks and finalizer modules
• ensure only a single beacon committees update can be run at any time
• optimize caching of beacon committees during block processing (thanks to @ariskk)
• optimize fetching of validator balances when calculating epoch summaries (thanks to @henridf)
• avoid crash when execution node does not return transaction for which it supplied a log entry

0.6.10

• avoid crash with uninitialised metrics
• avoid double fetching of blocks

0.6.8

• add missing column

0.6.7

• attempt to fetch chain configuration prior to genesis

0.6.6

• initialize monitoring metrics on startup to avoid drop to 0 on restart

0.6.5

• store validator epoch summaries individually if batch store fails

0.6.4

• fix fetching of execution paylods with blocks

0.6.3

• add f_timestamp to execution payload
• update go version to 1.18

0.6.2

• update to chain database libraries (no binary release)

0.6.1

• fix issue in obtaining sync committees after restart

0.6.0

• Bellatrix support

0.5.6

• do not fetch validator data twice for the same epoch

0.5.5

• fix bad index for sync aggregates

0.5.3

• correct summary value for attestations in epoch
• batch finality updates, reducing memory usage and increasing responsiveness
• avoid issue of duplicate column when upgrading f_parent_distance

0.5.1

• fix missed sync committee information if first block of period is missed

0.5.0

• add f_parent_distance to t_block_summaries
• fix crash on database disconnect
• use bulk insert for validator balances and validator epoch summaries

0.4.2:

• add index to t_deposits for validator public key

0.4.1:

• provide a more nuanced method for connecting to the chain database other than simple URL

0.4.0:

• Altair support

0.3.6:

• on failure to gather data wait and retry rather than mark as missed

0.3.5:

• wait until chain has started and node is synced before fetching data

0.3.4:

• update summarizer metadata for each update
• avoid crash in situation where transaction receipt is no longer available
• add HTTP request headers to avoid issue with latest version of geth
• update block metadata when attempt to obtain block is unsuccessful

0.3.3:

• avoid crash in situation where beacon committee information is not provided when expected
• update go-eth2-client dependency

0.3.2:

• standardise HTTP connections; avoid leaving connections open when no longer needed

0.3.1:

• fix issue with creation of database schema on clean installation

0.3.0:

• add summarizer module, providing summarized information for validators, blocks and epochs
• add t_epoch_summaries, t_block_summaries and t_validator_epoch_summaries tables
• add Ethereum 1 deposits module, queries an Ethereum 1 instance for deposits
• add section on querying the database, and notes on the individual tables
• fix potential crash in blocks handler if database goes away
• add prometheus metrics; details in docs/prometheus.md

0.2.0:

• automatically create initial tables if the database does not contain them
• add f_canonical, f_target_correct and f_head_correct to t_attestations; NULL if not yet sure
• add f_canonical to t_blocks to state if a block is canonical (true), non-canonical (false) or not yet sure (NULL)
• add provider functions to fetch aggregate validator balances
• add f_aggregation_indices to t_attestations to provide explicit indices for attestations
• add t_eth1_deposits table to hold data about deposits on the Ethereum 1 chain (N.B. not currently populated)
• add f_block_1_root and f_block_2_root to t_proposer_slashings table
• make log level configuration hierarchical; if a log level is not present it will use the next one up the hierarchy
• add t_chain_spec table, holding the chain specification
• add t_genesis table, holding the chain genesis information
• add t_deposits table, holding Ethereum 1 deposits recognized on the beacon chain and provided in the beacon block
• t_validators table no longer uses '-1' as an indicator for the Ethereum spec value FAR_FUTURE_EPOCH, instead it uses NULL
• add upgrader to update the database schema when changes are made