docs: update documentation and tests for disk cleanup solution

- Add comprehensive cleanup_disk_space.py documentation to maintenance README
- Document new Fly.io cleanup Makefile commands in Makefile.md
- Update test expectations to account for new cleanup job/schedule
- Include usage examples and safety features for all cleanup tools

Provides complete documentation for the disk space management
solution including automated and manual cleanup options.

Author: andrewm4894

Makefile.md

@@ -562,6 +562,43 @@ make posthog-example
 make kill-long-runs
 ```
 
+### Fly.io Disk Space Management
+
+#### `make fly-cleanup-preview`
+**Preview disk cleanup on Fly instance (dry run)**
+- Shows what files would be removed
+- Safe way to check cleanup impact
+- Requires `FLY_APP` environment variable
+
+```bash
+export FLY_APP=anomstack-demo
+make fly-cleanup-preview
+```
+
+#### `make fly-cleanup`
+**Clean up disk space on Fly instance**
+- Removes old artifacts (6+ hours)
+- Removes old logs (24+ hours)
+- Cleans database and runs VACUUM
+- Reports disk usage before/after
+
+```bash
+export FLY_APP=anomstack-demo
+make fly-cleanup
+```
+
+#### `make fly-cleanup-aggressive`
+**Emergency disk cleanup (aggressive mode)**
+- Removes artifacts older than 1 hour
+- Removes ALL log files
+- Use only when disk is critically full
+- More thorough than normal cleanup
+
+```bash
+export FLY_APP=anomstack-demo
+make fly-cleanup-aggressive
+```
+
 ### Legacy Targets
 
 #### `make docker-dev-env`

scripts/maintenance/README.md

@@ -39,6 +39,50 @@ python kill_long_running_tasks.py
 - **Validation**: Checks job status before taking action
 - **Error Handling**: Handles unreachable user code servers gracefully
 
+### `cleanup_disk_space.py`
+Standalone script for managing disk space by cleaning up old artifacts, logs, and metrics.
+
+**Features:**
+- **Artifact Cleanup**: Removes old Dagster run artifacts
+- **Log Cleanup**: Removes old log files from multiple directories
+- **Database Cleanup**: Removes old metrics and vacuums database
+- **Disk Usage Reporting**: Shows before/after disk usage statistics
+- **Dry Run Mode**: Preview cleanup without making changes
+- **Aggressive Mode**: More thorough cleanup for emergency situations
+
+**Use Cases:**
+- **Emergency Cleanup**: Free disk space when volume is full
+- **Scheduled Maintenance**: Regular cleanup to prevent disk issues
+- **Deployment Optimization**: Optimize Fly.io volume usage
+- **Development**: Clean up after testing
+
+**Usage:**
+```bash
+# Preview what would be cleaned up
+python cleanup_disk_space.py --dry-run
+
+# Normal cleanup (6h artifacts, 24h logs)
+python cleanup_disk_space.py
+
+# Aggressive cleanup (1h artifacts, all logs)
+python cleanup_disk_space.py --aggressive
+
+# Emergency cleanup with preview
+python cleanup_disk_space.py --dry-run --aggressive
+```
+
+**Cleanup Targets:**
+- **Artifacts**: Dagster run artifacts older than 6 hours (1 hour in aggressive mode)
+- **Logs**: Log files older than 24 hours (all logs in aggressive mode)
+- **Database**: Metrics older than 90 days + VACUUM operation
+- **Locations**: `/data/artifacts`, `/tmp/dagster`, `/data/dagster_storage`
+
+**Safety Features:**
+- **Dry Run Mode**: Safe preview of cleanup actions
+- **Detailed Reporting**: Shows exactly what will be/was removed
+- **Error Handling**: Continues cleanup even if individual files fail
+- **Size Calculation**: Reports space freed by cleanup operations
+
 ## Common Maintenance Tasks
 
 ### Regular Cleanup Operations

tests/test_main.py

@@ -9,19 +9,19 @@
 
 
 def test_jobs_len():
-    assert len(jobs) == 185
+    assert len(jobs) == 186  # Updated to include cleanup job
 
 
 def test_jobs_len_ingest():
-    assert len(ingest_jobs) == (len(jobs)-1) / 8
+    assert len(ingest_jobs) == (len(jobs)-2) / 8  # Updated to account for cleanup job (2 non-metric jobs total)
 
 
 def test_schedules_len():
-    assert len(schedules) == 185
+    assert len(schedules) == 186  # Updated to include cleanup schedule
 
 
 def test_schedules_len_ingest():
-    assert len(ingest_schedules) == (len(schedules)-1) / 8
+    assert len(ingest_schedules) == (len(schedules)-2) / 8  # Updated to account for cleanup schedule
 
 
 def test_jobs_schedules_len_match():