Bubble offers a dedicated hosting plan, where we host your apps on infrastructure isolated from other Bubble customers. Like with our shared plans, we run your apps: no need for you to deal with setting up servers and installing software, or monitoring them to keep them healthy. Unlike our shared plans, however, we run your app on dedicated servers that aren’t shared with any other customers, giving you increased reliability and the ability to scale as needed.
A typical dedicated installation consists of a database, a cache server, and one or more application servers. On a dedicated plan, we don’t monitor the number of workflows you run; instead, your only limit is the physical capacity of the hardware. We monitor the performance of your infrastructure, and if it looks like your app is slowing down due to more users than you have the hardware for (congratulations!), we’ll reach out to you and suggest an upgrade.
Dedicated hardware has the following benefits:
Scalability. We can quickly add more capacity to your dedicated cluster to deal with increased demand. We eventually hope to make this self-service so that you can scale up and down your own infrastructure.
More consistent application performance. Because you aren’t sharing the hardware with other Bubble customers, the only impact on application performance is how active your users are.
Greater reliability. Having your own infrastructure isolates you from the majority of Bubble-wide issue.
Data isolation. Your apps’ data is isolated from other customers’ data, leading to a stronger security profile.
Control. You decide when new Bubble code is rolled out to your cluster. Planning a major client meeting? You can wait until afterward to deploy the latest version of Bubble.
Regulatory compliance. If you need to host your app outside of the United States, this is possible under a dedicated plan. Please contact us for more details.
We work hard to keep your data safe. Some key points:
Our dedicated infrastructure is hosted on top of Amazon Web Services, which is a world-class hosting platform with extremely high security standards
User data on a dedicated plan is stored in a dedicated AWS RDS instance. Data is encrypted at rest, and the credentials for accessing the database are controlled by an automated system that logs all access to them.
Dedicated plans come with free SSL certificates to ensure that all communication with your app servers are secure. Once you have a certificate installed, we force all communication with the app to occur over https.
Bubble’s privacy rules are a powerful way of implementing application-level security. They provide you with fine-grained control over which items and which fields each user on your app can see and edit.
We offer point-in-time recovery for application changes and user data, meaning that you can jump back to how your data was at any point on the timeline. It’s up to you how long we retain data for -- you can scale up your data storage as needed to retain it for as long as you want. In addition, we use Amazon’s RDS automatic backups to ensure that we can recover from the total loss of a database server.
Bubble is new, rapidly developing technology. As such, we can’t yet recommend it for use cases where brief outages have disastrous business implications. That said, we are constantly working to keep your applications online. We monitor dedicated infrastructure 24 / 7, and respond as fast as possible to outages. We plan on offering uptime SLAs as the technology matures.
Dedicated plans offer two uptime advantages over shared hosting. The first is that, since your apps are isolated from other customers, issues with another customer’s app can’t spill over to affect your apps. The second is that we deploy new code to our shared hosting environment before deploying it to dedicated clusters, so you are less likely to be impacted by new bugs as we add features and extend the Bubble platform.
Finally, for customers who are especially concerned with uptime, we offer high-availability dedicated configurations. This means we have spare servers constantly running that can be swapped in if something goes wrong with a primary server. This is an expensive option (see the pricing section below), and the reliability difference between high availability dedicated and normal dedicated is smaller than the reliability difference between normal dedicated and shared hosting, so we don’t recommend this to all customers. In terms of greatest impact, having the database server as high-availability is most important, followed by the cache cluster, followed by the application server.