paradigmxyz
reth
Blog
Docs
Changelog
Blog
Docs
Changelog
Overview
Branches
Benchmarks
Runs
Performance History
Latest Results
bump
Soubhik-10:bal-devnet-2
40 minutes ago
refactor: use debug_assert for sorted targets check in batch pruning Amp-Thread-ID: https://ampcode.com/threads/T-019c275e-a92f-729b-905f-57ac2fce55c5
dan/history-pruning-perf
49 minutes ago
test: add batch pruning tests and debug assertions Amp-Thread-ID: https://ampcode.com/threads/T-019c2741-0710-763d-9bd1-90d9072f03b1
dan/history-pruning-perf
1 hour ago
chore: fix clippy warnings and clean up comments Amp-Thread-ID: https://ampcode.com/threads/T-019c2741-b7cf-746e-9fb9-4ab5641e81ea
yc/unwind-history-batch
1 hour ago
test: add batch pruning tests and debug assertions Amp-Thread-ID: https://ampcode.com/threads/T-019c2736-8282-7154-98c7-d3d4b5cd37f8
dan/history-pruning-perf
1 hour ago
perf(rocksdb): use single iterator for batch history unwind Previously, unwind_account_history_indices and unwind_storage_history_indices created a new iterator_cf for every address/storage key by calling account_history_shards/storage_history_shards in a loop. This is the same performance antipattern fixed in PR #21767 for pruning. This commit adds: - unwind_account_history_batch: batch version using single raw_iterator_cf - unwind_storage_history_batch: batch version using single raw_iterator_cf - RocksDBRawIterEnum: wrapper supporting seek() for iterator reuse The batch methods reuse a single raw iterator and skip seeks when the iterator is already positioned correctly (for sorted targets in key order). This significantly reduces RocksDB seek overhead for large unwind operations. Amp-Thread-ID: https://ampcode.com/threads/T-019c272c-eed5-751b-a0c4-8e260f96a3bc
yc/unwind-history-batch
2 hours ago
fix
Soubhik-10:trace-bal
2 hours ago
fix
Soubhik-10:trace-bal
2 hours ago
Active Branches
chore: update fixture for bal test
last run
40 minutes ago
#21787
CodSpeed Performance Gauge
-11%
perf(rocksdb): use single iterator for batch history unwind
last run
1 hour ago
#21786
CodSpeed Performance Gauge
0%
fix(prune): report HasMoreData when limiter breaks segment loop early
last run
2 hours ago
#21785
CodSpeed Performance Gauge
0%
© 2026 CodSpeed Technology
Home
Terms
Privacy
Docs