Add automated PostgreSQL performance index management

- Add scripts and SQL for base and supplementary indexes (total 23)
- Integrate index checks and auto-creation into Jellyfin startup
- Update csproj to include SQL files in build/publish output
- Add diagnostics, publish/copy scripts, and troubleshooting docs
- Provide detailed guides for migration, verification, and merging
- Ensures robust, automated, and well-documented index setup
This commit is contained in:
2026-02-28 15:13:04 -05:00
parent 79110d2afd
commit 5565dc3e05
35 changed files with 5469 additions and 0 deletions
+126
View File
@@ -0,0 +1,126 @@
# Quick Connect and Add Supplementary Indexes
This script connects to your `jellyfin_testsdata` database and adds the 5 supplementary indexes.
## Prerequisites
1. PostgreSQL is running
2. Database `jellyfin_testsdata` exists
3. `psql` is in your PATH
4. You have the jellyfin user credentials
## Quick Run
### Option 1: Direct SQL (Fastest)
```powershell
# Just add the indexes directly
psql -U jellyfin -d jellyfin_testsdata -f sql\schema_init\10_create_supplementary_indexes.sql
```
### Option 2: With verification and checks
```powershell
# Run the full script with checks and verification
.\scripts\Apply-SupplementaryIndexes.ps1
```
### Option 3: Manual Step-by-Step
1. **Connect to database**:
```powershell
psql -U jellyfin -d jellyfin_testsdata
```
2. **Check current indexes**:
```sql
SELECT indexname
FROM pg_indexes
WHERE schemaname = 'library'
AND indexname LIKE 'idx_baseitems_%'
ORDER BY indexname;
```
3. **Apply supplementary indexes**:
```sql
\i sql/schema_init/10_create_supplementary_indexes.sql
```
4. **Verify creation**:
```sql
SELECT
indexname,
pg_size_pretty(pg_relation_size(indexrelid)) as size,
idx_scan as times_used
FROM pg_stat_user_indexes
WHERE indexname IN (
'idx_baseitems_type_isvirtualitem_topparentid',
'idx_baseitems_topparentid_isfolder',
'idx_baseitems_datecreated_filtered',
'idx_itemvaluesmap_itemvalueid_itemid',
'idx_activitylogs_userid_datecreated'
);
```
## What Gets Added
5 supplementary indexes:
1. `idx_baseitems_type_isvirtualitem_topparentid` - Filtered library browsing
2. `idx_baseitems_topparentid_isfolder` - Folder hierarchy
3. `idx_baseitems_datecreated_filtered` - Recently added view
4. `idx_itemvaluesmap_itemvalueid_itemid` - Genre/tag reverse lookup
5. `idx_activitylogs_userid_datecreated` - User activity queries
## Troubleshooting
### "CONCURRENTLY cannot be used" error
Remove `CONCURRENTLY` from the SQL if you get this error (less common in newer PostgreSQL versions).
### "Index already exists"
This is safe - the script checks before creating. The index already exists and doesn't need to be created again.
### "Connection refused"
Ensure PostgreSQL is running:
```powershell
# Check if PostgreSQL service is running
Get-Service postgresql*
# Or try to ping the database
psql -U jellyfin -d jellyfin_testsdata -c "SELECT version();"
```
### Wrong password
Update the password in the script or use `.pgpass` file:
```powershell
# Windows: %APPDATA%\postgresql\pgpass.conf
# Linux/Mac: ~/.pgpass
# Format: hostname:port:database:username:password
localhost:5432:jellyfin_testsdata:jellyfin:YOUR_PASSWORD
```
## Monitor Performance
After applying, monitor index usage:
```powershell
# Wait a few days, then check usage
psql -U jellyfin -d jellyfin_testsdata -c "
SELECT
indexname,
idx_scan as times_used,
idx_tup_read as rows_read,
pg_size_pretty(pg_relation_size(indexrelid)) as size
FROM pg_stat_user_indexes
WHERE indexname LIKE 'idx_%'
AND schemaname = 'library'
ORDER BY idx_scan DESC;
"
```
## Rollback
If you need to remove the indexes:
```sql
DROP INDEX CONCURRENTLY IF EXISTS library.idx_baseitems_type_isvirtualitem_topparentid;
DROP INDEX CONCURRENTLY IF EXISTS library.idx_baseitems_topparentid_isfolder;
DROP INDEX CONCURRENTLY IF EXISTS library.idx_baseitems_datecreated_filtered;
DROP INDEX CONCURRENTLY IF EXISTS library.idx_itemvaluesmap_itemvalueid_itemid;
DROP INDEX CONCURRENTLY IF EXISTS activitylog.idx_activitylogs_userid_datecreated;
```