What Is It?
MongoDB is a fully managed NoSQL document database service offering three editions: Playground (free single-instance for testing), Business (predefined templates), and Enterprise (fully customizable with sharding support). The service provides replica sets for high availability, automatic daily backups with 7-day retention, point-in-time recovery (Enterprise), and built-in monitoring via Grafana dashboards. Clusters run on private LANs with configurable resources up to 31 cores, 230 GB RAM, and 4 TB storage per instance.
Quick Facts
| Aspect | Details |
|---|---|
| Type | Managed NoSQL document database |
| Editions | Playground (free testing), Business (templates), Enterprise (custom + sharding) |
| Cluster Types | Replica Set (all editions), Sharded Cluster (Enterprise only, 2-32 shards via DCD, up to 100 via API) |
| Replication | Replica set with odd number of members, write-concern default: majority + j:true |
| Scaling | Existing clusters cannot be scaled in-place (requires new cluster or support request) |
| Monitoring | Prometheus + MongoDB exporter, visualized in Grafana |
| Network Access | Private LAN only (no direct internet exposure) |
Resource Limits by Edition
| Edition | CPU Cores | RAM | Storage | Nodes | Sharding |
|---|---|---|---|---|---|
| Playground | 1 vCPU | 2 GB | 50 GB | 1 node (single instance) | No |
| Business | Predefined templates | Predefined templates | Predefined templates | 1 or 3 nodes (replica set; only Business can downscale from 3 to 1) | No |
| Enterprise | 1-31 cores | Up to 230 GB | Up to 4 TB | 1, 3, 5, or 7 nodes (2-32 shards via DCD, up to 100 via API) | Yes |
Absolute Maximums
| Component | Maximum |
|---|---|
| Concurrent Connections | ~114,000 (at the maximum 230 GB RAM per instance; formula: (RAM in GB - 2) / 2, expressed in thousands) |
| CPU Cores per Instance | 31 |
| RAM per Instance | 230 GB |
| Storage per Instance | 4 TB |
| Shards (Enterprise) | 32 (DCD create) / 100 (via API) |
| Backup Retention | 7 days |
Edition Comparison
| Feature | Playground | Business | Enterprise |
|---|---|---|---|
| Cost | Free | Paid | Paid |
| Customizable Resources | No | No | Yes |
| Sharded Clusters | No | No | Yes |
| Backups | No | Yes (7 days) | Yes (7 days + PITR) |
| BI Connector | No | No | Yes (optional) |
| Point-in-Time Recovery | No | No | Yes (1-24h window) |
What You Can Do
Document-Based Storage
Store and query data as flexible JSON-like documents (BSON format) with dynamic schemas. Supports nested documents, arrays, and rich data types without requiring predefined table structures.
High Availability with Replica Sets
Deploy replica sets with an odd number of members (default 3 for Business/Enterprise). Primary node handles writes while secondaries replicate data and can serve reads. Automatic failover promotes a secondary to primary during failures.
Horizontal Scaling with Sharding
Create sharded clusters (Enterprise only) to distribute data across shards (2-32 via the DCD create flow, up to 100 by growing shards via the API). Three config-server instances are automatically provisioned (2 cores, 4 GB RAM, 40 GB storage each) at no additional cost. Choose shard key to partition data across the cluster.
Note: Sharding is only available in Enterprise edition.
Automated Backups and Recovery
Daily snapshots are automatically retained for 7 days (Business and Enterprise only). Restore clusters from any snapshot via DCD or API. Enterprise edition adds point-in-time recovery (PITR) with 1-24 hour window for restoring to specific timestamps.
Note: Playground edition does not include backup functionality.
Performance Monitoring
Built-in Grafana dashboards (via Logging Service) expose comprehensive metrics including CPU and memory usage, storage and disk I/O, connection counts, replication lag, operation counts (queries, inserts, updates, deletes), index usage, and read/write patterns.
BI Connector for SQL Queries
Enable MongoDB Connector for Business Intelligence (Enterprise only) to run SQL-style queries against MongoDB data. Integrates with standard BI tools and reporting platforms.
User and Role Management
Create database users via DCD UI with customizable usernames, passwords, built-in roles, and target databases. Users can be edited or deleted after creation. Role-based access control ensures fine-grained security.
WiredTiger Storage Engine
Automatic cache management allocates 50% of (RAM - 1 GB) to WiredTiger cache with minimum 256 MB. Each client connection consumes up to 1 MB RAM outside the cache. Oplog retained for 24 hours.
Read and Write Preferences
Configure write-concern and read-preference settings client-side. Default write-concern uses "majority" with j:true for durability. Read-preference can be set to primary, secondary, or nearest based on application requirements.
Best For
| Scenario | Why It Fits |
|---|---|
| Rapid prototyping and testing | Playground edition provides free single-instance cluster for development and experimentation |
| Web and mobile applications | Flexible document schemas adapt to evolving data models without migrations |
| Content management systems | Nested documents and arrays naturally represent hierarchical content structures |
| Real-time analytics | High read/write throughput with configurable read preferences for analytics workloads |
| Internet of Things (IoT) | Horizontal sharding (Enterprise) scales to handle massive data volumes from device fleets |
| Catalog and product databases | Dynamic schemas accommodate varying product attributes without schema changes |
| User profiles and session storage | Document model maps directly to application objects without object-relational impedance |
Key Considerations
Billing & Costs
- Main billing: Per hour based on edition, CPU cores, RAM, and storage provisioned per instance
- Playground edition: Free tier (no billing)
- Business edition: Billed according to selected template resources
- Enterprise edition: Billed for custom resources (CPU, RAM, storage) per node
- Additional costs: Config servers for sharded clusters are not billed; backup storage included in 7-day retention
Limitations
- Existing clusters cannot be scaled in-place (must create new cluster or request Enterprise scaling via IONOS support)
- Maximum 31 CPU cores, 230 GB RAM, and 4 TB storage per instance (subject to account quota)
- Maximum concurrent connections is approximately 114,000, reached only at the maximum 230 GB RAM per instance (formula: (RAM in GB - 2) / 2, expressed in thousands)
- Automatic backups retained for 7 days only
- Point-in-time recovery limited to 1-24 hour window (Enterprise only)
- Forbidden private IP ranges: 172.16.0.0/12, 192.168.230.0/24, 10.208.0.0/12
- Playground edition: single node only (no high availability), no backups, fixed resources (1 vCPU, 2 GB RAM, 50 GB storage)
- Business edition: cannot customize resources (must use predefined templates), no sharding support
- Sharded clusters and BI Connector available only in Enterprise edition
- WiredTiger cache uses approximately 50% of available RAM (affects application memory allocation)
- Storage must be sized for data, indexes, journal, and the Oplog (retaining at least 24 hours of write history), plus a 20-30% growth buffer; access logs are shipped off-volume to central storage with a 30-day retention and do not consume the database instance's disk space
Management Options
- IONOS Data Center Designer (DCD) web interface for cluster creation and management
- IONOS DBaaS REST API for programmatic control
- MongoDB Shell (mongosh) for database operations and queries
- MongoDB Compass GUI for visual database exploration
- Standard MongoDB drivers for application integration
- Grafana dashboards (via Logging Service) for monitoring and metrics visualization
- Prometheus metrics for custom monitoring integrations