Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge bitcoin#28280: Don't empty dbcache on prune flushes: >30% faste…
…r IBD 589db87 validation: don't erase coins cache on prune flushes (Andrew Toth) 0e89187 Add linked-list test to CCoinsViewCache::SanityCheck (Pieter Wuille) 05cf4e1 coins: move Sync logic to CoinsViewCacheCursor (Andrew Toth) 7825b8b coins: pass linked list of flagged entries to BatchWrite (Andrew Toth) a14edad test: add cache entry linked list tests (Andrew Toth) 24ce37c coins: track flagged cache entries in linked list (Andrew Toth) 58b7ed1 coins: call ClearFlags in CCoinsCacheEntry destructor (Andrew Toth) 8bd3959 refactor: require self and sentinel parameters for AddFlags (Andrew Toth) 75f36d2 refactor: add CoinsCachePair alias (Andrew Toth) f08faea refactor: move flags to private uint8_t and rename to m_flags (Andrew Toth) 4e4fb4c refactor: disallow setting flags in CCoinsCacheEntry constructors (Andrew Toth) 8737c0c refactor: encapsulate flags setting with AddFlags and ClearFlags (Andrew Toth) 9715d3b refactor: encapsulate flags get access for all other checks (Andrew Toth) df34a94 refactor: encapsulate flags access for dirty and fresh checks (Andrew Toth) Pull request description: Since bitcoin#17487 we no longer need to clear the coins cache when syncing to disk. A warm coins cache significantly speeds up block connection, and only needs to be fully flushed when nearing the `dbcache` limit. For frequent pruning flushes there's no need to empty the cache and kill connect block speed. However, simply using `Sync` in place of `Flush` actually slows down a pruned full IBD with a high `dbcache` value. This is because as the cache grows, sync takes longer since every coin in the cache is scanned to check if it's dirty. For frequent prune flushes and a large cache this constant scanning starts to really slow IBD down, and just emptying the cache on every prune becomes faster. To fix this, we can add two pointers to each cache entry and construct a doubly linked list of dirty entries. We can then only iterate through all dirty entries on each `Sync`, and simply clear the pointers after. With this approach a full IBD with `dbcache=16384` and `prune=550` was 32% faster than master. For default `dbcache=450` speedup was ~9%. All benchmarks were run with `stopatheight=800000`. | | prune | dbcache | time | max RSS | speedup | |-----------:|----------:|------------:|--------:|-------------:|--------------:| | master | 550 | 16384 | 8:52:57 | 2,417,464k | - | | branch | 550 | 16384 | 6:01:00 | 16,216,736k | 32% | | branch | 550 | 450 | 8:05:08 | 2,818,072k | 8.8% | | master | 10000 | 5000 | 8:19:59 | 2,962,752k | - | | branch | 10000 | 5000| 5:56:39 | 6,179,764k | 28.8% | | master | 0 | 16384 | 4:51:53 | 14,726,408k | - | | branch | 0 | 16384 | 4:43:11 | 16,526,348k | 2.7% | | master | 0 | 450 | 7:08:07 | 3,005,892k | - | | branch | 0 | 450 | 6:57:24 | 3,013,556k |2.6%| While the 2 pointers add memory to each cache entry, it did not slow down IBD. For non-pruned IBD results were similar for this branch and master. When I performed the initial IBD, the full UTXO set could be held in memory when using the max `dbcache` value. For non-pruned IBD with max `dbcache` to tip ended up using 12% more memory, but it was also 2.7% faster somehow. For smaller `dbcache` values the `dbcache` limit is respected so does not consume more memory, and the potentially more frequent flushes were not significant enough to cause any slowdown. For reviewers, the commits in order do the following: First 4 commits encapsulate all accesses to `flags` on cache entries, and then the 5th makes `flags` private. Commits `refactor: add CoinsCachePair alias` to `coins: call ClearFlags in CCoinsCacheEntry destructor` create the linked list head nodes and cache entry self references and pass them into `AddFlags`. Commit `coins: track flagged cache entries in linked list` actually adds the entries into a linked list when they are flagged DIRTY or FRESH and removes them from the linked list when they are destroyed or the flags are cleared manually. However, the linked list is not yet used anywhere. Commit `test: add cache entry linked list tests` adds unit tests for the linked list. Commit `coins: pass linked list of flagged entries to BatchWrite` uses the linked list to iterate through DIRTY entries instead of using the entire coins cache. Commit `validation: don't erase coins cache on prune flushes` uses `Sync` instead of `Flush` for pruning flushes, so the cache is no longer cleared. Inspired by [this comment](bitcoin#15265 (comment)). Fixes bitcoin#11315. ACKs for top commit: paplorinc: ACK 589db87 sipa: reACK 589db87 achow101: ACK 589db87 mzumsande: re-ACK 589db87 Tree-SHA512: 23b2bc01c83edacb5b39aa60bb0b766de9a74ce17f0c59bf13b97b4328a7b758ad9aff6581c3ca88e2973f7658380651530d497444f48d6e22ea0bfc51cc921d
- Loading branch information