fix: probe actual GIS capability instead of inferring it from the DB platform - #30
Open
torvalstrom wants to merge 1 commit into
Open
fix: probe actual GIS capability instead of inferring it from the DB platform#30torvalstrom wants to merge 1 commit into
torvalstrom wants to merge 1 commit into
Conversation
…platform A PostgreSQL platform does not imply PostGIS. On a plain postgres server (the default for most Nextcloud installs) detectGisType() returned GIS_TYPE_POSTGRES and every queryPoint() call ran PostGIS SQL that can only fail: one logged error and one doomed query per photo per cron run (~8.6k log lines/day on a ~30k-photo library) before falling back. Probe capability instead of assuming it: - postgres: check pg_extension for 'postgis' - mysql/mariadb: try a trivial ST_Contains() (built-in since MySQL 5.7 / MariaDB 10.2, but a stripped build should also degrade cleanly) When the probe fails, detectGisType() returns GIS_TYPE_NONE so queryPoint() goes straight to fallbackByFileId(), which resolves places from the oc_memories_places table Memories precomputes at index time - place names still resolve fully, with zero errors. Verified on Nextcloud 33 + plain postgres 16: occ journeys:cluster over 64k photos, 1105 clusters, 0 resolver errors (previously ~47k/day).
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
SimplePlaceResolver::detectGisType()infers GIS capability from the DB platform name: any PostgreSQL platform is assumed to have PostGIS. On a plain postgres server (no PostGIS extension - the common case for Nextcloud installs) everyqueryPoint()call then runs spatial SQL that can only fail:error_logline per photo per cron run - on a ~30k-photo library that was ~8,600 log lines/day, and ~47k/day on a 64k-photo libraryFix
Probe the actual capability once, in the constructor path:
SELECT 1 FROM pg_extension WHERE extname = 'postgis'ST_Contains()(built-in since MySQL 5.7 / MariaDB 10.2, but a stripped build also degrades cleanly this way)If the probe fails,
detectGisType()returnsGIS_TYPE_NONE, soqueryPoint()goes straight tofallbackByFileId()- which resolves places from theoc_memories_placestable that Memories precomputes at index time. Place names still resolve fully; all three callers already passfileid.Verified
Running this exact change in production on Nextcloud 33 + plain postgres 16 since July:
occ journeys:clusterover 64,341 photos → 1,105 clusters, 0 resolver errors (previously tens of thousands per day). Behavior on a server that does have PostGIS is unchanged (probe returns true, spatial path used as before).🤖 Generated with Claude Code