diff options
| author | Marian Beermann <public@enkore.de> | 2017-05-26 12:30:15 +0200 |
|---|---|---|
| committer | Marian Beermann <public@enkore.de> | 2017-06-02 17:43:14 +0200 |
| commit | 9f8b967a6f45bb7dbbf1f37cf231ea82f149a0f6 (patch) | |
| tree | 5b81c5b941968e8366d4a1de7cb32a947853553d | |
| parent | 740898d83ba9d83589bcbd607050d23007de318a (diff) | |
cache sync: initialize master index to known capacity
| -rw-r--r-- | src/borg/cache.py | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/src/borg/cache.py b/src/borg/cache.py index c9fa70b7f..cd3a9951d 100644 --- a/src/borg/cache.py +++ b/src/borg/cache.py @@ -603,6 +603,9 @@ Chunk index: {0.total_unique_chunks:20d} {0.total_chunks:20d}""" # deallocates old hashindex, creates empty hashindex: chunk_idx.clear() cleanup_outdated(cached_ids - archive_ids) + # Explicitly set the initial hash table capacity to avoid performance issues + # due to hash table "resonance". + master_index_capacity = int(len(self.repository) / ChunkIndex.MAX_LOAD_FACTOR) if archive_ids: chunk_idx = None if self.progress: @@ -630,7 +633,7 @@ Chunk index: {0.total_unique_chunks:20d} {0.total_chunks:20d}""" # Do not make this an else branch; the FileIntegrityError exception handler # above can remove *archive_id* from *cached_ids*. logger.info('Fetching and building archive index for %s ...', archive_name) - archive_chunk_idx = ChunkIndex() + archive_chunk_idx = ChunkIndex(master_index_capacity) fetch_and_build_idx(archive_id, repository, self.key, archive_chunk_idx) logger.info("Merging into master chunks index ...") if chunk_idx is None: @@ -641,7 +644,7 @@ Chunk index: {0.total_unique_chunks:20d} {0.total_chunks:20d}""" else: chunk_idx.merge(archive_chunk_idx) else: - chunk_idx = chunk_idx or ChunkIndex() + chunk_idx = chunk_idx or ChunkIndex(master_index_capacity) logger.info('Fetching archive index for %s ...', archive_name) fetch_and_build_idx(archive_id, repository, self.key, chunk_idx) if self.progress: |