Platform
High availability
High availability (HA) keeps your project online when the server it runs on fails. If a node goes down, Lofenta moves your project to a healthy node automatically. You do not need to do anything. Available on the Developer plan and up.
What it protects against
HA protects against a single runtime server going down, for example a hardware fault, a kernel panic, or a network outage on that node. Lofenta detects that the node has stopped responding and brings your project back up on another healthy node. Your data comes with it, because Lofenta keeps a copy of your project on a second node ahead of time.
Standard vs Premium
Standard
Included
When the node fails, Lofenta rebuilds your project on a healthy node and routes traffic back to it. The rebuild takes a few minutes. There is a short interruption while it happens, and no data is lost.
- Recovery time. A few minutes.
- During recovery. Visitors see a transfer page, not an error.
- Cost. Included on Developer and up.
Premium
Paid add-on
Lofenta keeps a second copy of your project running and warm on another node at all times. When the main node fails, traffic switches to the warm copy inside a single request. There is no transfer page and effectively no downtime.
- Recovery time. Effectively instant.
- During recovery. Visitors notice nothing.
- Cost. Paid, because it runs a second copy around the clock.
Premium runs a full second instance of your project on another node at all times, so it is priced to cover that second instance. Standard only uses a second node at the moment of a failure, so it is included.
What visitors see during a failover
On Standard, while your project is being rebuilt on a healthy node, anyone who opens your app sees a Lofenta transfer page that reads Transferring to another node. It explains that the project is moving and will be back shortly. The page clears itself the moment your project is live again. No refresh or action is needed.
That page carries internal code 660. You can look it up alongside every other platform code in the internal error codes reference.
On Premium there is no transfer page. The switch happens at the network layer inside one request, so visitors keep browsing as if nothing happened.
Your data
Lofenta keeps a copy of your project files and managed databases on a second node ahead of a failure, so when your project moves it brings its data with it. There is a small replication window, which means writes made in the final seconds before a sudden node loss may not have reached the copy yet. For most workloads this gap is negligible. If you need stronger guarantees, talk to support about your setup.