516 Commits (aa3d5b5e5fab8d7910f6383dc0bcb80a77ace8d2)

Author	SHA1	Message	Date
Ayrat Badykov	00f12015cc	remove pending ops on block invalidation	5 years ago
Ayrat Badykov	b92ad5e9ce	fix merging conflicts	5 years ago
Ayrat Badykov	2bd6f66e99	remove nonconsensus blocks from pending blocks	5 years ago
pasqu4le	9f08c1c373	test fix, comments and small refactorings	5 years ago
pasqu4le	6a933d514f	internal_txs	5 years ago
Ayrat Badykov	99cb7972ae	add block_hash to token_transfers	5 years ago
Ayrat Badykov	187f0dc402	fix realtime fetcher small skips feature ... If indexing is started from scratch, realtime fetcher tries to index large ranges of blocks. For example, 5..1852044. It produces errors like `failed to fetch: :emfile. Block will be retried by catchup indexer` This PR limits the number of blocks to 10	5 years ago
Ayrat Badykov	5fddd3ded8	add last_block_number=0 for empty node and db	5 years ago
Ayrat Badykov	496f8f7907	fix logging level	5 years ago
Ayrat Badykov	a2d947c4c8	log token instances fetching to a separate file	5 years ago
pasqu4le	e8c2fa554e	Allow Realtime Fetcher to wait for small skips ... Problem: Sometimes the realtime fetcher's `newHeads` websocket subscription skips a small amount of block numbers, just to be followed by those very same numbers that were skipped, without repeating the number that caused the skipping. This may be cause by out-of-order events or be a legitimate reorg, but in any case this causes close-in-time refetching which in turn leads to async insertions problems. Solution: allow the Realtime fetcher to keep track of a small (settable) skip and wait for the block number inside the skip to arrive before fetching the whole window. Obviously if it gets notified of the blocks following the skip (or it is forced into polling) the whole window gets fetched and inserted.	5 years ago
Ayrat Badykov	1fefec100f	save no uri error to instance	5 years ago
Ayrat Badykov	483beaa10d	do not fetch token instances with errors	5 years ago
pasqu4le	f04052c4c4	Remove nonconsensus blocks from cache after internal transactions importing ... Problem: after PR #2726 some blocks may lose consensus after internal transactions are imported. The cache is not aware of this and may hold outdated values. Solution: notify the cache to drop the blocks that lost consensus in such cases.	5 years ago
Ayrat Badykov	e644949b8d	do not make heavy db request	5 years ago
Ayrat Badykov	a73082a5fc	add error logging	5 years ago
Ayrat Badykov	d34275d2fe	on-fly fetching of token instances ... Currently, we have to restart indexer to fetch metadata for newly imported ERC-721 token instances. This PR triggers fetching of new token instances after blocks imports in Catchup and Realtime fetchers.	5 years ago
Ayrat Badykov	0a168dec6d	add TokenUpdater to supervision tree	5 years ago
Ayrat Badykov	487acfa643	fix dialyzer	5 years ago
Ayrat Badykov	e59c41144e	fix credo	5 years ago
Ayrat Badykov	de497d8d94	fix token updating	5 years ago
pasqu4le	c83c0d39c7	Add cache for first page of uncles	5 years ago
Ayrat Badykov	1676189c3f	fix build	5 years ago
Ayrat Badykov	a30e65fa94	rewrite token updater	5 years ago
Ayrat Badykov	156c6ca455	do not update cache if no blocks were inserted	5 years ago
pasqu4le	624946a12b	Add pending transactions cache	5 years ago
Ayrat Badykov	818cf823bb	convert smart contract hash to string	5 years ago
pasqu4le	af7782f2fe	Add ETS-based cache for accounts page	5 years ago
pasqu4le	a8cd5058cf	Enforce DB transaction's order to prevent deadlocks ... Problem: due to the ShareLock mechanism used by PostgreSQL when there are several DB transactions are acting on multiple rows of the same table, it's possible to incur in a deadlock and so into an error. Solution: enforce a consistent order of ShareLock acquisition in all the relevant transaction. Also add clear documentation that should help in the future.	5 years ago
Ayrat Badykov	c8db2a5f50	fix token instance fetcher	5 years ago
Ayrat Badykov	5c12d9e3cb	add TokenInstance Fetcher	5 years ago
Yegor San	772bc92e6c	Changes staging	5 years ago
pasqu4le	3f30610dde	Add generic Map-like Cache behaviour and implementation ... Problem: multiple caches exist that contain some data as a map (key-value storing). More are to be implemented, so it would be good to abstract their implementation. Solution: add a behaviour and a macro (providing an implementation) to create this type of caches. At the same time, where possible: - redefine the existing ones using the new macro - optimize the implementation	5 years ago
Ayrat Badykov	0e736637a7	skip cache update if there are no blocks inserted ... Sometimes due to some error (mostly network) if may not fetch any data when indexing block. If block cache update is triggered it fails. This PR check if list is passed to cache update function.	5 years ago
pasqu4le	b52d10b22a	Add generic Ordered Cache behaviour and implementation ... Problem: multiple caches exist to keep a page of the most recent `n` elements. More are to be implemented, so it would be good to abstract their implementation. Solution: add a behaviour and a macro (providing an implementation) to create this type of caches. Redefine the existing ones using the new macro.	5 years ago
Ayrat Badykov	5361c4e824	fix logging for finished indexed with last block set	5 years ago
Ayrat Badykov	8f92a1f720	fix cr issues	5 years ago
saneery	12ebc6fad8	Add LAST_BLOCK var to reindex range of blocks	5 years ago
Ayrat Badykov	c360439be9	fix complexity credo warning	5 years ago
Ayrat Badykov	d426b2f758	fix remaining case warnings	5 years ago
saneery	1e4de6e41e	Add env to disable indexer supervisor	5 years ago
pasqu4le	da775b07b7	Group Explorer caches ... Problem: a bunch of different caches are under the Explorer.Chain namespaces, cluttering the codebase Solution: create a subdirectory and collect them all under the Explorer.Chain.Cache namespace	5 years ago
pasqu4le	54a1010a11	Improve Address controllers ... This is yet another step in checking controllers for improvements. In particular this time Address controllers were checked and this (mostly refactoring) changes were made: - a few calls to `Explorer.Chain.hash_to_address` were modified to avoid useless preloads where possible - `Explorer.Chain.find_or_insert_address_from_hash` was modified for the very same reason, because it calls the previous function - `Explorer.Chain.address_to_logs` has been modified to have a more legible where clause - both `Explorer.Chain.hash_to_address` and `Explorer.Chain.token_from_address_hash` were modified to support `necessity_by_association_option`s for consistency	5 years ago
Victor Baranov	9bbec7b2cd	Remove changes, that fixed on Nethermind side	6 years ago
pasqu4le	bef66e846f	Add transactions cache ... Problem: due to the unsatisfying performance of the main page some ETS-based cache has been deemed necessary. In this particular case one for the most recent collated transactions. Solution: implementation of said transactions cache	6 years ago
pasqu4le	144f39db5b	Remove duplicate params in `Indexer.Fetcher.TokenBalance` ... Problem: `Explorer.Chain.Import.Runner.Address.TokenBalances` fails with the `cardinality_violation` error because it tries to insert more than once token balances (for the same {token_contract_address_hash, address_hash, block_number} tuple) in a single database transaction. Solution: filter out duplicate parameters in `Indexer.Fetcher.TokenBalance` so that no duplicates are sent to be imported.	6 years ago
pasqu4le	920cfe45bb	Reduce BlocksTransactionsMismatch memory footprint	6 years ago
Ayrat Badykov	857ac30058	use seconds	6 years ago
pasqu4le	ebb1628976	Improve token's controllers by avoiding unnecessary preloads	6 years ago
Ayrat Badykov	b594c3c8af	fix more Nethermind errors	6 years ago