fix(files): clean up filecache companion rows of removed storages - #64208
Merged
susnux merged 1 commit intoSep 17, 2026
Merged
Conversation
cleanByMountId() deleted the filecache rows of an unmounted storage but left their filecache_extended and file metadata rows behind. It now goes through Storage::removeFileCacheEntries(), the same path clear() uses. files:cleanup also removes filecache_extended, files_metadata and files_metadata_index rows whose file is no longer in the filecache, so rows orphaned before this fix are cleaned up too. Signed-off-by: Git'Fellow <12234510+solracsf@users.noreply.github.com>
CarlSchwan
approved these changes
Sep 11, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Deleting an external storage removed its filecache rows through
Storage::cleanByMountId()but left theirfilecache_extended,files_metadataandfiles_metadata_indexrows behind. They now go through the same batched helper asCache::clear().occ files:cleanupnow also removes rows in those three tables whose file id is no longer in the filecache, so instances that already have orphans can clean them up. This is a full walk of each table (1000 file ids per batch, across all shards) instead of only looking at deleted storages, so it takes longer on large instances.--skip-filecache-extendedstill skips thefilecache_extendedpass, and the command prints two extra lines for the metadata tables.The new tests fail without the change and pass with it on MariaDB 11.8 and PostgreSQL 16 (local and S3 primary storage) and on a 4-shard MariaDB setup.
Checklist
3. to review, feature component)stable32)AI (if applicable)