Platform availability & reliability
Quickbase uses several measures to achieve high reliability and high availability.
Redundancy of the runtime environment
Quickbase maintains production-ready compute capacity for all core infrastructure components within each location.
Runtime or application failures
If a runtime instance gets into an unhealthy state:
- The control plane automatically detects the issue.
- Quickbase automatically migrates applications pinned to that instance to other runtime instances.
- Customers do not need to take action here, and they will not even be aware if such a migration occurs.
Disaster recovery
Quickbase maintains multiple geographically diverse hosting regions. All necessary data is continuously replicated across regions. In the event of one region's complete failure, we can quickly switch over and run from another region instead. We routinely switch between hosting locations multiple times per year to exercise our switchover capability.
The details behind our approach to disaster recovery depend on the hosting platform:
- For Amazon AWS, Quickbase uses cloud architecture best practices We also design our services to run active/active across multiple availability zones within a region.
- For Google GCP, Quickbase uses cloud architecture best practices. We also design our services to run in a high availability configuration.
To better understand Quickbase’s RPO and other relevant details, please visit the Quickbase Trust Center.
Reliability and uptime
We designed Quickbase with enterprise scale in mind, and we have an average uptime of 99.9%. Quickbase makes uptime stats widely available through our service page. Customers are encouraged to monitor this page to see news of upcoming maintenance, or other platform service updates. You can also use the subscribe option on the service page to get email alerts.
Load testing
Quickbase creates redundancy and ensures that the platform is available. Quickbase builders work with tools in the platform and with Quickbase Solution Architects to make sure individual applications perform well and are available to their users. With Quickbase’s purpose-built, in-memory database, you can measure and anticipate performance by considering request complexity, table data size, and the number of requests in a given time window. Because of this, traditional load testing is not supported.
If you try to run a traditional load test on your Quickbase application, you may see false successes or failures that do not represent real-world usage. Synthetic tests also may not represent the variety of users and dynamic permissions.
Quickbase provides numerous tools to help builders assess and improve application performance. These tools not only surface actionable insights but can also automatically apply performance optimizations safely to production applications.
The best way to achieve high confidence is through small, controlled roll-outs with all roles represented. Then, work with Quickbase Solution Architects to determine scale ceilings and strategies based on data gathered.