PostgreSQL by VSHN

This is a preview - the Service is currently in development.

PostgreSQL, managed by VSHN for you.

PostgreSQL is a powerful, open source object-relational database system that uses and extends the SQL language combined with many features that safely store and scale the most complicated data workloads. The origins of PostgreSQL date back to 1986 as part of the POSTGRES project at the University of California at Berkeley and has more than 30 years of active development on the core platform.

— PostgreSQL Website

Features

Standalone

Availability

Number of replicas

1 primary replica

No downtime during Maintenance

Multiple logical databases per instance

Multiple users per instance

Super-user access

Maximum supported size of database

500 GiB [1]

Guaranteed performance

[2]

High availability

Support

(see Support Plans)

More architectures will be documented once available.

The following common features are available as well:

Self-Service Instance Provisioning and Configuration

Service instances can be self-service ordered and are provisioned fully automated in the background. Access to self-service provisioning is provided via custom Kubernetes configuration objects in the Kubernetes cluster or via Open Service Broker API (on request).

Best-Practice Configuration Management

Services are configured with best practices and the configuration is updated continuously as we learn improvements during day-2 operations and from the community. Industry standards are used for security configuration and wherever possible TLS encrypted connections are offered by default.

Maintenance and Security Operations

Patch updates are applied as they get available. We monitor security information for services and apply zero-day patches or workaround configuration as they become available.

Data Protection and Recovery

All services offer a regular backup option. See the service descriptions for more details.

Service Metrics

Performance metrics are collected regularly and are available as graphs to the user. The metrics are automatically monitored and acted upon irregularities, see Monitoring and Alerting below.

Monitoring and Alerting

Service Level Indicators are monitored and alerted upon. Depending on the service level alerts are automatically handled by a VSHNeer.

Service Logging

All logs are collected and are available to debug service errors.

Support by VSHN or Vendor

Support for all services are available from VSHN support engineers. Depending on the service, additional support by the software vendor or a third-party might be available.

Supported versions

Supported versions:

See Versioning Policy of the StakGres project to read about more detailed versions support. Once a version isn’t supported anymore by the upstream project, we’ll stop supporting the version as well.

Upgrade Policy

  • We only support a version as long as it’s supported upstream

  • 3 months before the support ends ("Final Release") the grace period starts, and you’re urged to migrate to a supported version

  • Once the support for a service has ended, all managed services are stopped, but the service still continues to run, but without any guarantee

  • We currently do not offer an in-place upgrade of Major versions. To switch to a newer Major version (for example from version 13 to version 14) a new instance has to be deployed, and the data has to be migrated manually.

SLIs

SLA relevant

According to the service levels the SLI "Up" is defined as follows:

Condition "Up"
  • PostgreSQL responds to 95% of "select 1" requests within a 1-minute interval. The requests are originating from within the same cluster.

Condition "Down"
  • PostgreSQL is not responding to "select 1" requests within a 1-minute interval.

Incident Prevention

The following SLIs are defined to prevent incidents:

  • The configured storage volume is 90% full.

  • The remaining available connections are less than 10% over 30 minutes

  • The remaining memory available is less than 10% over 30 minutes

Data Protection

  • Backups can be configured per instance

  • By default backups happen at 22:00 each day

  • By default 6 backups are kept

  • The schedule and retention is configurable

  • Backups are not encrypted at rest

Recurring Maintenance

Planned Maintenance Activities
  • Update the base container image with the latest patched version, causes service restart

  • Only security updates are part of the automated maintenance. Updates to database versions which need additional work (for example code update, data migration) is a separate task and has to be taken care of by the end user.

  • Configuration improvements as we learn new best-practices or some configuration proved to be non-optimal

Mandatory Maintenance Activities
  • Security issue (CVE) mitigation by updating to a patched version or configuration adjustments, usually causes service restart

  • Regular maintenance on the underlying platform, usually causing up to 2 restarts per replica

Maintenance Windows
  • Automated once a week

  • Day and time can be configured per service instance

  • Default is a random time in the night from Tuesday to Wednesday

Users responsibilities

As a DevOps company, we believe in its collaborative approach. Flawless service is only possible through a sense of responsibility on both sides. Accordingly, we rely on the user to consider the following points:

  • Choose reasonably sized resources for the software using the service

  • Act early when an increase of service usage is foreseen (for example increase resources)

  • Rely on best-practices for using the service

  • Choose a matching service level for your use-case

Behind the scenes

This service is provided by StackGres. Service instances are running on the same Kubernetes cluster as the service instance was ordered in.

Pricing

Architecture Service Level

Best Effort

Guaranteed Availability

Converged Standalone

CHF 0.1666666667 per hour
CHF 120.00 per 30 days (720h)

CHF 0.5 per hour
CHF 360.00 per 30 days (720h)

  • All prices are per instance

  • Excl. Compute resources

  • Validity: These prices are valid starting 2022-XX-XX until further notice.

See Pricing for more details.

Available PostgreSQL extensions

Please check StackGres' list of extensions.

Alternatively a list is available directly in the database with select * from pg_available_extensions;. To use an extension in your database, consult the documentation on CREATE EXTENSION — install an extension.


1. Cumulated over all logical databases
2. depends on the underlying platform