sombra 0.3.2

High-performance graph database with ACID transactions, single-file storage, and bindings for Rust, TypeScript, and Python
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
# Sombra Production Deployment Guide

This guide provides comprehensive instructions for deploying Sombra v0.2.0 in production environments.

## Table of Contents

1. [Hardware Requirements]#hardware-requirements
2. [Operating System Configuration]#operating-system-configuration
3. [Filesystem Recommendations]#filesystem-recommendations
4. [Application Configuration]#application-configuration
5. [Monitoring and Observability]#monitoring-and-observability
6. [Backup and Recovery]#backup-and-recovery
7. [High Availability Patterns]#high-availability-patterns
8. [Security Considerations]#security-considerations
9. [Performance Tuning]#performance-tuning
10. [Troubleshooting]#troubleshooting

## Hardware Requirements

### Minimum Requirements

**Small Deployments** (< 1M nodes, < 10M edges):
- CPU: 2 cores
- RAM: 512MB + (cache_size × 8KB)
- Disk: 10GB SSD
- I/O: 100 IOPS

**Medium Deployments** (< 10M nodes, < 100M edges):
- CPU: 4 cores
- RAM: 2GB + (cache_size × 8KB)
- Disk: 100GB SSD
- I/O: 500 IOPS

**Large Deployments** (< 100M nodes, < 1B edges):
- CPU: 8+ cores
- RAM: 8GB + (cache_size × 8KB)
- Disk: 1TB NVMe SSD
- I/O: 5000+ IOPS

### Recommended Specifications

For production workloads, we recommend:

```
CPU: 4-8 cores (for parallel operations)
RAM: 4-16GB (generous cache sizing)
Disk: SSD or NVMe (NVMe 3-5x faster for WAL)
Network: 1Gbps+ (for monitoring/backup)
```

### Storage Sizing

**Formula:**
```
Total Storage = (Database Size × 1.5) + WAL Overhead + Backup Space

Where:
  Database Size = estimated graph size
  WAL Overhead = max_wal_size_mb × 2 (safety margin)
  Backup Space = Database Size × retention_days
```

**Example:**
```
Database: 100GB graph
WAL: 500MB × 2 = 1GB
Backups: 100GB × 7 days = 700GB
Total: 100GB + 1GB + 700GB = 801GB
```

## Operating System Configuration

### Linux (Recommended)

#### File Descriptor Limits

Sombra requires file descriptors for the database file, WAL, and monitoring:

```bash
# /etc/security/limits.conf
* soft nofile 65536
* hard nofile 65536

# Verify
ulimit -n
# Should show: 65536
```

#### Virtual Memory Tuning

```bash
# /etc/sysctl.conf

# Reduce swappiness (prefer RAM over swap)
vm.swappiness=10

# Increase dirty page writeback time for better batching
vm.dirty_writeback_centisecs=500
vm.dirty_expire_centisecs=3000

# Apply changes
sudo sysctl -p
```

#### I/O Scheduler

For SSDs, use `noop` or `deadline` scheduler:

```bash
# Check current scheduler
cat /sys/block/sda/queue/scheduler

# Set to deadline (better for databases)
echo deadline > /sys/block/sda/queue/scheduler

# Make persistent (add to /etc/rc.local or systemd)
```

For NVMe drives, use `none`:

```bash
echo none > /sys/block/nvme0n1/queue/scheduler
```

#### Transparent Huge Pages

Disable THP for predictable performance:

```bash
# Temporary
echo never > /sys/kernel/mm/transparent_hugepage/enabled
echo never > /sys/kernel/mm/transparent_hugepage/defrag

# Permanent (add to /etc/rc.local)
```

### macOS

macOS generally has good default settings, but consider:

```bash
# Increase file descriptor limit
ulimit -n 65536

# Add to ~/.zshrc or ~/.bash_profile
echo "ulimit -n 65536" >> ~/.zshrc
```

### Windows

```powershell
# Run as Administrator

# Disable Windows Defender real-time scanning for database directory
Add-MpPreference -ExclusionPath "C:\path\to\sombra\data"

# Disable search indexing for database files
# Use GUI: File Explorer → Right-click folder → Properties → 
#         Uncheck "Allow files in this folder to have contents indexed"
```

## Filesystem Recommendations

### Linux Filesystems

**ext4** (Recommended for most deployments):
- Pros: Mature, stable, good performance
- Cons: Slightly slower than XFS for large files
- Configuration:
  ```bash
  # Mount with noatime for better performance
  mount -o noatime,nodiratime /dev/sda1 /data
  
  # Add to /etc/fstab:
  /dev/sda1  /data  ext4  noatime,nodiratime  0  2
  ```

**XFS** (Recommended for large databases):
- Pros: Better performance for large files, good parallelism
- Cons: Slightly more complex
- Configuration:
  ```bash
  # Format with larger allocation groups
  mkfs.xfs -d agcount=16 /dev/sda1
  
  # Mount with optimal options
  mount -o noatime,nodiratime,logbufs=8,logbsize=256k /dev/sda1 /data
  ```

**Btrfs** (Use with caution):
- Pros: Snapshots, compression
- Cons: Potential performance issues, less mature
- Not recommended for production Sombra deployments

### macOS Filesystems

**APFS** (Default, recommended):
- Good SSD performance
- Native macOS support
- No special configuration needed

### Windows Filesystems

**NTFS**:
- Disable 8.3 filename generation:
  ```powershell
  fsutil 8dot3name set 1
  ```
- Defragmentation: Not needed for SSDs

## Application Configuration

### Production Configuration Template

```rust
use sombra::db::config::Config;

let config = Config {
    // Core settings
    page_size: 8192,                    // 8KB pages (standard)
    cache_size: 5000,                   // 40MB cache (adjust based on RAM)
    enable_wal: true,                   // Always true in production
    
    // Resource limits
    max_database_size_mb: Some(50_000), // 50GB limit (prevent disk exhaustion)
    max_wal_size_mb: 500,               // 500MB WAL (auto-checkpoint trigger)
    max_transaction_pages: 10_000,      // Limit transaction size
    
    // Timeouts
    transaction_timeout_ms: Some(60_000), // 1 minute timeout
    auto_checkpoint_interval_ms: Some(30_000), // 30 second auto-checkpoint
    
    // Safety
    fsync_enabled: true,                // Full durability
    checksum_mode: true,                // Enable checksums (when available)
};

let db = GraphDB::open_with_config("/data/sombra/graph.db", config)?;
```

### Configuration by Workload

#### Read-Heavy Workload
```rust
Config {
    cache_size: 10_000,              // Large cache (80MB)
    max_wal_size_mb: 100,            // Small WAL (less writes)
    auto_checkpoint_interval_ms: Some(60_000), // Less frequent checkpoints
    ..Config::production()
}
```

#### Write-Heavy Workload
```rust
Config {
    cache_size: 2_000,               // Moderate cache (16MB)
    max_wal_size_mb: 1_000,          // Large WAL buffer
    auto_checkpoint_interval_ms: Some(10_000), // Frequent checkpoints
    max_transaction_pages: 50_000,   // Large transactions
    ..Config::production()
}
```

#### Balanced Workload
```rust
Config::production()  // Use defaults
```

### Environment-Specific Settings

```rust
// Development
let config = if cfg!(debug_assertions) {
    Config {
        cache_size: 100,
        enable_wal: false,  // Faster for tests
        ..Config::default()
    }
// Staging
} else if env::var("ENV").unwrap_or_default() == "staging" {
    Config {
        cache_size: 1000,
        max_wal_size_mb: 100,
        ..Config::production()
    }
// Production
} else {
    Config::production()
};
```

## Monitoring and Observability

### Structured Logging Setup

```rust
use sombra::logging::init_logging;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Initialize logging based on environment
    let log_level = env::var("LOG_LEVEL").unwrap_or_else(|_| "info".to_string());
    init_logging(&log_level)?;
    
    // Your application code
    let db = GraphDB::open("production.db")?;
    
    Ok(())
}
```

**Log Levels in Production:**
- `error`: Always enable (critical failures)
- `warn`: Always enable (degradation, slow queries)
- `info`: Recommended (transactions, checkpoints) - < 2% overhead
- `debug`: Only for troubleshooting (3-5% overhead)
- `trace`: Never in production (10-15% overhead)

### Metrics Collection

```rust
use std::time::Duration;
use std::thread;

// Metrics reporting thread
thread::spawn(move || {
    loop {
        thread::sleep(Duration::from_secs(60));
        
        let metrics = db.metrics();
        
        // Export to monitoring system
        report_to_prometheus(&metrics);
        
        // Log summary
        info!(
            "Metrics: tx_commit={} tx_rollback={} cache_hit_rate={:.2}% wal_bytes={}",
            metrics.transactions_committed,
            metrics.transactions_rolled_back,
            metrics.cache_hit_rate() * 100.0,
            metrics.wal_bytes_written
        );
    }
});
```

### Prometheus Integration

```rust
// Export metrics in Prometheus format
fn report_to_prometheus(metrics: &PerformanceMetrics) {
    let prometheus_text = metrics.to_prometheus_format();
    
    // Write to file for node_exporter textfile collector
    std::fs::write("/var/lib/node_exporter/sombra.prom", prometheus_text)?;
    
    // Or expose via HTTP endpoint
    // prometheus_server.update_metrics(prometheus_text);
}
```

**Prometheus Queries:**
```promql
# Transaction rate
rate(sombra_transactions_committed_total[5m])

# Error rate
rate(sombra_corruption_errors_total[5m])

# Cache hit rate
sombra_cache_hits / (sombra_cache_hits + sombra_cache_misses) * 100

# P95 commit latency
histogram_quantile(0.95, sombra_commit_latency_seconds)
```

### Health Checks

```rust
use std::time::Duration;

// Health check endpoint (for load balancers, orchestrators)
fn health_check(db: &GraphDB) -> Result<HealthStatus> {
    let health = db.health_check();
    
    match health.status {
        HealthStatus::Healthy => {
            info!("Health check: HEALTHY");
            Ok(health.status)
        },
        HealthStatus::Degraded => {
            warn!("Health check: DEGRADED - {}", 
                  health.checks.iter()
                      .filter(|c| !c.healthy)
                      .map(|c| c.description())
                      .collect::<Vec<_>>()
                      .join(", "));
            Ok(health.status)
        },
        HealthStatus::Unhealthy => {
            error!("Health check: UNHEALTHY - {}", 
                   health.checks.iter()
                       .filter(|c| !c.healthy)
                       .map(|c| c.description())
                       .collect::<Vec<_>>()
                       .join(", "));
            Err(GraphError::DatabaseUnhealthy)
        }
    }
}
```

### Alerting Rules

**Critical Alerts** (Page on-call):
- Database corruption detected
- Lock poisoning (requires restart)
- Disk space < 10% free
- Error rate > 1% of requests
- Service down for > 1 minute

**Warning Alerts** (Slack/email):
- Cache hit rate < 70%
- WAL size > 80% of max
- P95 latency > 10ms
- Transaction timeout rate > 0.1%
- Health check degraded

**Info Alerts** (Dashboard only):
- Checkpoint completed
- WAL rotated
- Cache size changed

## Backup and Recovery

### Backup Strategies

#### 1. Snapshot Backups (Recommended)

```bash
#!/bin/bash
# backup.sh - Daily database backup

DATE=$(date +%Y%m%d)
DB_PATH="/data/sombra/graph.db"
BACKUP_DIR="/backups/sombra"
RETENTION_DAYS=7

# Trigger checkpoint (flushes WAL)
/usr/local/bin/sombra-repair checkpoint "$DB_PATH"

# Copy database file
cp "$DB_PATH" "$BACKUP_DIR/graph-$DATE.db"

# Compress
gzip "$BACKUP_DIR/graph-$DATE.db"

# Remove old backups
find "$BACKUP_DIR" -name "graph-*.db.gz" -mtime +$RETENTION_DAYS -delete

echo "Backup completed: graph-$DATE.db.gz"
```

**Schedule with cron:**
```cron
# Daily at 2 AM
0 2 * * * /usr/local/bin/backup.sh >> /var/log/sombra-backup.log 2>&1
```

#### 2. Continuous Backup (WAL Archiving)

```bash
#!/bin/bash
# wal-archive.sh - Continuous WAL backup

WAL_DIR="/data/sombra"
ARCHIVE_DIR="/backups/sombra/wal-archive"

# Watch for WAL rotations and copy
inotifywait -m -e close_write "$WAL_DIR" |
while read -r directory events filename; do
    if [[ "$filename" == *"-wal" ]]; then
        cp "$directory/$filename" "$ARCHIVE_DIR/"
        echo "Archived: $filename"
    fi
done
```

#### 3. Cloud Backup (S3 Example)

```bash
#!/bin/bash
# s3-backup.sh - Upload backups to S3

BACKUP_FILE="/backups/sombra/graph-$(date +%Y%m%d).db.gz"
S3_BUCKET="s3://my-company-backups/sombra"

# Upload to S3
aws s3 cp "$BACKUP_FILE" "$S3_BUCKET/" --storage-class STANDARD_IA

# Verify
aws s3 ls "$S3_BUCKET/" | grep "$(date +%Y%m%d)"
```

### Recovery Procedures

#### 1. Point-in-Time Recovery (from snapshot)

```bash
# 1. Stop application
systemctl stop myapp

# 2. Restore database file
gunzip -c /backups/sombra/graph-20251020.db.gz > /data/sombra/graph.db

# 3. Remove WAL (will be rebuilt)
rm -f /data/sombra/graph.db-wal
rm -f /data/sombra/graph.db-shm

# 4. Verify integrity
sombra-inspect verify /data/sombra/graph.db

# 5. Start application
systemctl start myapp
```

#### 2. Recovery from Corruption

```bash
# 1. Assess damage
sombra-inspect verify /data/sombra/graph.db

# 2. Attempt repair
sombra-repair checkpoint /data/sombra/graph.db

# 3. If repair fails, restore from backup
# (see Point-in-Time Recovery above)

# 4. Report corruption details for analysis
sombra-inspect stats /data/sombra/graph.db > corruption-report.txt
```

#### 3. Disaster Recovery (total data loss)

```bash
# 1. Provision new infrastructure
# 2. Restore latest backup
# 3. Restore application state
# 4. Validate data integrity
# 5. Resume operations

# See: docs/disaster-recovery-runbook.md (create this for your org)
```

### Backup Testing

**Monthly Backup Validation:**
```bash
#!/bin/bash
# test-backup.sh - Validate backups are restorable

LATEST_BACKUP=$(ls -t /backups/sombra/graph-*.db.gz | head -1)
TEST_DIR="/tmp/sombra-backup-test"

mkdir -p "$TEST_DIR"
gunzip -c "$LATEST_BACKUP" > "$TEST_DIR/graph.db"

# Verify
sombra-inspect verify "$TEST_DIR/graph.db"
RESULT=$?

rm -rf "$TEST_DIR"

if [ $RESULT -eq 0 ]; then
    echo "✓ Backup valid: $LATEST_BACKUP"
    exit 0
else
    echo "✗ Backup corrupted: $LATEST_BACKUP"
    exit 1
fi
```

## High Availability Patterns

### 1. Active-Passive Failover

```
┌─────────────┐         ┌─────────────┐
│   Primary   │◄────────│  Secondary  │
│   (Active)  │ Backup  │  (Standby)  │
└─────────────┘         └─────────────┘
       │                       │
       ▼                       ▼
   Clients              (ready to serve)
```

**Implementation:**
- Primary serves all traffic
- Secondary receives continuous backups (or replication in future version)
- On failure, promote secondary to primary
- DNS/load balancer failover

**Failover procedure:**
```bash
# Detect failure
if ! health_check primary; then
    # Promote secondary
    systemctl stop myapp@primary
    systemctl start myapp@secondary
    
    # Update DNS
    update-dns-record myapp.example.com -> secondary-ip
    
    # Alert ops team
    send-alert "Primary failed, secondary promoted"
fi
```

### 2. Read Replicas (Future v0.3.0+)

*Note: Not available in v0.2.0. Requires MVCC implementation.*

### 3. Application-Level Sharding

```rust
// Partition graph by entity type
fn get_shard(entity_id: u64) -> usize {
    (entity_id % NUM_SHARDS) as usize
}

let shards: Vec<GraphDB> = (0..NUM_SHARDS)
    .map(|i| GraphDB::open(format!("/data/shard-{}.db", i)))
    .collect::<Result<Vec<_>>>()?;

// Route operations to appropriate shard
let shard = get_shard(node_id);
shards[shard].get_node(node_id)?;
```

## Security Considerations

### File Permissions

```bash
# Database files should be owned by application user only
chown myapp:myapp /data/sombra/graph.db
chmod 600 /data/sombra/graph.db

# WAL files
chmod 600 /data/sombra/graph.db-wal
chmod 600 /data/sombra/graph.db-shm
```

### Encryption at Rest

Sombra does not provide built-in encryption. Use filesystem-level encryption:

**Linux (LUKS):**
```bash
# Encrypt partition
cryptsetup luksFormat /dev/sda1
cryptsetup open /dev/sda1 sombra-data

# Mount encrypted volume
mount /dev/mapper/sombra-data /data
```

**macOS (FileVault):**
```bash
# Enable FileVault via System Preferences → Security & Privacy
```

### Network Security

- **No network exposure**: Sombra is an embedded database (no network protocol)
- **Application-level auth**: Implement authentication in your application
- **API security**: Use HTTPS, API keys, OAuth, etc. for your application API

### Audit Logging

```rust
// Log all mutations for audit trail
fn add_node_with_audit(db: &mut GraphDB, node: Node, user: &str) -> Result<NodeId> {
    let node_id = db.add_node(node)?;
    
    audit_log!(
        user = user,
        action = "add_node",
        node_id = node_id,
        timestamp = Utc::now()
    );
    
    Ok(node_id)
}
```

## Performance Tuning

See [docs/performance.md](performance.md) for detailed performance analysis and tuning guide.

**Quick tips:**
- Cache size = (working set size / page_size)
- Batch operations in transactions (100-1000 ops/tx)
- Use property indexes for frequent queries
- Monitor cache hit rate (target: >80%)
- Checkpoint every 1000-10000 transactions

## Troubleshooting

### Common Issues

#### 1. High Memory Usage

**Symptoms:** Process memory grows over time

**Diagnosis:**
```rust
let metrics = db.metrics();
println!("Cache size: {}MB", (metrics.cache_size * 8192) / 1_000_000);
println!("Page evictions: {}", metrics.page_evictions);
```

**Solutions:**
- Reduce `cache_size` in config
- Check for memory leaks in application code
- Monitor with `valgrind` or `heaptrack`

#### 2. Slow Performance

**Symptoms:** High latency, low throughput

**Diagnosis:**
```rust
let metrics = db.metrics();
println!("Cache hit rate: {:.2}%", metrics.cache_hit_rate() * 100.0);
println!("P95 commit latency: {}ms", metrics.p95_commit_latency());
```

**Solutions:**
- Increase `cache_size` (if RAM available)
- Batch small operations into transactions
- Create property indexes for frequent queries
- Use SSD/NVMe storage

#### 3. Database Corruption

**Symptoms:** `GraphError::Corruption` errors

**Diagnosis:**
```bash
sombra-inspect verify /data/sombra/graph.db
```

**Solutions:**
- Restore from latest backup
- Check disk health (SMART errors)
- Ensure `fsync_enabled = true` in production
- Enable checksums (when available)

#### 4. Lock Poisoning

**Symptoms:** `GraphError::LockPoisoned` errors

**Cause:** Another thread panicked while holding a lock

**Solution:**
```bash
# Requires application restart
systemctl restart myapp
```

**Prevention:**
- Fix any panic paths in application code
- Use `std::panic::catch_unwind` for external code
- Monitor for panics in logs

### Getting Support

- **Documentation**: [https://docs.rs/sombra]https://docs.rs/sombra
- **GitHub Issues**: [https://github.com/maskdotdev/sombra/issues]https://github.com/maskdotdev/sombra/issues
- **Examples**: Check the `examples/` directory

## Production Readiness Checklist

Before deploying to production:

- [ ] Hardware meets minimum requirements
- [ ] OS tuning applied (file descriptors, vm.swappiness, I/O scheduler)
- [ ] Filesystem configured (noatime, proper scheduler)
- [ ] Production configuration reviewed (`Config::production()`)
- [ ] Structured logging enabled (INFO level)
- [ ] Metrics collection configured
- [ ] Health checks implemented
- [ ] Monitoring dashboards created
- [ ] Alerting rules configured (critical + warning)
- [ ] Backup strategy implemented and tested
- [ ] Backup restoration tested monthly
- [ ] Disaster recovery plan documented
- [ ] Security hardening applied (file permissions, encryption)
- [ ] Performance baseline established
- [ ] Capacity planning completed
- [ ] Runbooks created for common issues
- [ ] On-call rotation trained
- [ ] Graceful shutdown implemented (`db.close()`)

---

**Congratulations! You're ready to run Sombra in production.** 🚀

For questions or issues, please file a GitHub issue or check the documentation at [docs.rs/sombra](https://docs.rs/sombra).