observability See every block of every slow request.
A visual backend is only as good as its debugging story. Fluxify traces a route block by block, keeps every failed run, and sends it all to the tools you already use.
Read the docssampling
Keep every failure. Sample the rest.
Tracing is switched on per route, because detail has a cost. What's exported is decided after each run ends: failed runs are always kept, successful ones are sampled.
destinations
Your stack, not ours.
Logs, traces and metrics each get their own destination: Loki for logs, any OpenTelemetry collector for all three. Telemetry leaves the worker off the request path, so a slow collector never slows a caller.
in the editor
Logs where you can find them.
Logging blocks and the script logger write to the project's log destination. Compile errors and other system events show up in the editor next to the route they belong to.
-
Console Log
print to the server logs from any point in a flow
-
Cloud Logs
ship a line straight to Loki or OTLP
-
Per-route tracing
on for the routes you're watching, off for the hot ones
-
Compile status
a failed save is reported on the route, with the reason