PostgreSQL

Description

PostgreSQL is a powerful, open-source object-relational database system that has been under active development and improvement for over 30 years. It has earned a strong reputation for reliability, robust functionality, and performance. We use the original software from the PostgreSQL Global Development Group for core functionality. We install and operate your database servers, ensuring optimal configuration in coordination with connected systems. Our monitoring service tracks the system and proactively alerts you to issues such as overload or potential data loss. We optimize cache sizes to deliver that extra boost in speed. Databases and their integration can be complex, depending on your specific requirements. If you are unsure which of our options best suits your application, please feel free to contact our consulting team. Our experts will help you meet your performance, availability, and resilience requirements. Full database backups are performed daily, separate from the Managed Cloud Instance backups. Our solution ensures consistent database backups. If a replica is present, the backup is created from it, ensuring the primary instance's performance remains unaffected. Retention periods are 7 days, 4 weeks, and 3 months. Our goals for this service:

High-performance, reliable database services
High configuration flexibility
Data storage in ISO 27001-certified data centers

What you can expect from our service

Consistent daily database backups
Scalable architecture
Monitoring of all relevant health and performance metrics
Up to 100 GB of storage for your database
Intelligent auto-configuration following any vCPU or RAM changes
Point-in-time recovery of a replica instance to the time of the last daily database backup

Optional service features

Additional DB instances
More than 100 GB of secondary storage

Service requests

Service restart
Performance scaling
Simple configuration changes
Service stop or uninstallation
Point-in-time recovery
Enabling built-in plugins
Disabling built-in plugins

Standard Configurations

We offer two distinct standard hosting solutions that differ in performance and availability. Database content—such as data, schemas, and indexes—is imported by the customer. If these options do not meet your requirements, or if you would like assistance with the setup, please contact our consulting team to develop a solution tailored to your specific needs.

Single-Server Setup

This option is particularly suitable for development and testing systems, as it offers excellent value for money. Only a single database service is installed and operated as the primary instance, handling both read and write operations. Adding a replica at a later stage is generally straightforward. However, the primary instance will experience a temporary increase in load while the initial backup is created to establish replication to the replica. Backups are performed during live operation and may impact service performance.

Primary/Replica Setup (Our Best Practice)

For production environments, we recommend a primary/replica setup that allows read and write operations to be handled separately. All write operations are directed to the primary instance, while a significant portion of read operations is handled by the replica (provided your client application supports this). This distributes the read load across multiple instances, resulting in higher performance. Read performance can be scaled without interruption by adding further replicas; if a primary instance fails, we ensure that a replica takes over the primary's duties. As a preventive measure, we always include replicas with an identical configuration unless expressly agreed otherwise. This is a requirement for the Premium SLA. All replicas synchronize automatically with the primary instance. We continuously monitor the time lag between changes on the primary and the replicas, taking action if a specific threshold is exceeded. Database backups are created by the replica, which offloads the primary instance. We offer an optional "SLA Premium" service for this configuration.

Our trending tool gives you access to historical data, such as:

Number of server processes
Number of "committed" and "rolled back" transactions
Process types and counts (running/sleeping/zombie) allowing you to monitor and optimize your client application's use of the database servers.

External Access

We do not grant access to database services from the public internet. Please contact our consulting team if you require this.

Optional Features/Configurations

Distributing the DB Setup Across Multiple Availability Zones

Except for the single-server setup, any configuration variant can be distributed across multiple availability zones. This ensures the service remains available—or can be quickly restored—even if a complete data center failure occurs.

Managed Cloud Instance Sizes

Upon request, we would be happy to provide a quote for larger or smaller managed cloud instances tailored to your performance requirements. High-performance storage options are also available.

Discounts for Non-Production Servers

Many customers also host their development (DEV) and test (STAGE) systems with us. Since these systems typically have lower performance and availability requirements than production systems, we can use the underlying hardware more densely. Scheduling maintenance tasks—such as reboots or migrations—is also less complex. This allows us to offer a discount for staging and development systems. Should you wish to switch such an environment to production-grade infrastructure, we can arrange this upon request; please note that the discount specified in your hosting contract would no longer apply in this case. Please contact our sales team if you have any questions, and refer to the relevant sections of your SLA agreement for these systems.

FAQ

Can the primary and replica be located in different availability zones?

Yes, that is possible.

What happens if the database disk space runs low?

Our alerting system triggers an alarm as the disk space limit approaches. We attempt to increase the storage capacity temporarily in the short term; however, it is your responsibility to stay within the disk storage limits. Please contact our Service Management team to receive a custom quote for servers requiring more than 100 GB of permanent storage.

Can the primary and replica fail simultaneously?

We distribute the primary and replica across different hardware to achieve greater component redundancy.

Is there a performance drop during database backups?

The specific impact depends on the chosen setup. In a primary/replica setup, only the replica's performance is affected.

Important Information

We offer PostgreSQL as a managed service. You can find further information and details about our managed services on our website.