4d83b687d4
feat: Add metrics for file read triggered
...
Usually, "Loading block data from ..." shouldn't be shown in logs at all. Add metrics to detect the file read.
2025-10-05 11:28:11 +00:00
12f366573e
fix: Do not increase block counter when no block is read
...
This made ingest loop to infinitely increase the block number
2025-10-05 11:28:11 +00:00
b8bae7cde9
fix: Utillize LruMap better
...
LruMap was introduced to allow getting the same block twice, so removing the item when getting the block doesn't make sense.
2025-10-05 11:28:11 +00:00
0fd4b7943f
refactor: Use offsets instead of lines, wrap related structs in one
2025-10-05 11:28:04 +00:00
aa73fab281
chore: Now cargo fmt sorts imports and trait methods
2025-10-05 03:56:23 -04:00
66c2ee654c
feat: Add block source metrics
2025-09-11 18:50:22 -04:00
dd2c925af2
fix: Sort hl-node files correctly
2025-08-26 13:47:34 -04:00
52909eea3f
feat: Add --local.fallback-threshold, --s3.polling-interval
2025-08-25 22:27:26 -04:00
65cdc27b51
fix: line_to_evm_block don't hold equivalent semantic after refactor
2025-08-24 16:46:45 +08:00
19f35a6b54
chore: clippy, fmt
2025-08-24 01:15:36 -04:00
d61020e996
refactor: Split files for block sources
...
By claude code
2025-08-24 01:14:33 -04:00
73a34a4bc1
chore: clippy
2025-08-23 22:17:03 -04:00
d8eef6305b
remove: Reduce unnecessary LoC
2025-08-23 22:17:03 -04:00
bae68ef8db
refactor: Reduce unnecessary LoC
...
By claude code
2025-08-23 04:21:23 -04:00
4be1aa83de
Port to reth 1.6.0-dev
2025-08-21 05:58:37 -04:00
239ee5f8e8
remove: Remove logs that degrades performance
...
This happens when syncing from genesis while local-ingest-dir exists. Find better way of logging them, and before that, disable it.
2025-08-08 02:06:45 -04:00
bd9a0020e6
refactor: Code style, lint
2025-08-04 02:50:39 -04:00
a8df1fdaeb
fix: Reduce fallback request before even next block is mined
...
Currently it was doing 0.5s + (0.25s x N); now it's 5s + (0.25s x N), assuming 5 block behind is bad enough to request fallback. Also usually fallback "exists" so that it updates last poll time acordingly, so it won't poll much after fallback is triggered.
2025-08-02 23:57:51 -04:00
c27e5e5a64
fix/perf: Fix last line scanner, wait 0.5s before fallback if it's more recent, add tests
2025-08-02 17:10:30 -04:00
c0b3acf181
perf: Reduce log
2025-08-01 17:37:29 +00:00
77158aa164
perf: Do not allocate much when backfilling ranges
2025-08-01 17:30:33 +00:00
2d6b5e5cd2
chore: Improve log
2025-08-01 17:28:31 +00:00
ff2e55b5a2
perf: Do not use cutoff when backfilling
2025-08-01 17:20:09 +00:00
0180711ae4
perf: Constrain memory size again, add log
2025-07-31 01:56:39 -04:00
a766ee0236
feat: Support range-based backfill for hl-node ingestion
2025-07-31 01:51:22 -04:00
46c9d4cbf9
fix: Fix path parser
2025-07-31 01:00:28 -04:00
78ae5643b1
fix: Should use fallback when there is no hl-node files
2025-07-30 23:32:25 -04:00
4e59ee62dc
fix: Reduce fallback usage in hl-node ingestion
2025-07-30 23:29:18 -04:00
c34ee01b70
fix: Do not use LruMap
...
LruMap does not support backfill.
2025-07-30 22:20:21 -04:00
9fcc04e889
fix: Use correct cutoff block number
2025-07-30 21:27:55 -04:00
a2e978dc0c
fix typo
2025-07-30 04:37:49 +00:00
a47674ac30
feat: Add backfill support from local node
...
This currently does not support huge amount of backfill log within local node
2025-07-30 04:05:52 +00:00
f605c56bdf
chore: cargo fix, fmt and misc fixes
2025-07-02 16:06:32 +00:00
d90ddbbc65
chore: more log
2025-07-02 10:45:24 +00:00
4b078a1b2d
chore: clippy
2025-07-01 03:19:02 +00:00
c3fb631ce4
doc: a little bit more docs for hl_node ingestion
2025-07-01 03:15:24 +00:00
1633b3d2e1
feat: hl-node ingestion
...
Ported from https://github.com/hl-archive-node/nanoreth/pull/7 ; since this repo is experimental, will be merged with nanoreth repo in future.
2025-07-01 03:09:03 +00:00
