Clusters left running overnight. Warehouses idling all weekend. Platform teams set allowed schedules once. Kostavo scans continuously and stops resources running outside their window.
A developer spins up an interactive cluster at 5pm and forgets to stop it. It runs all night and through the weekend.
Dev and test warehouses stay running outside business hours with no active queries: pure waste.
Pipeline triggers and job schedules firing at unexpected times, running workloads that weren't intended to run off-hours.
By the time anyone notices, days or weeks of after-hours compute have already run for nothing.
Platform teams define the allowed window for dev and test resources. Kostavo scans continuously. Anything running outside that window gets stopped and the resource owner is notified.
Real schedules from the product: business-hours protection on weekdays, 24/7 freeze for production.
Dev resources running 24/7 means 128 hours a week where no one is working, but you're still paying.
Detect and stop dev/test clusters running outside allowed hours
Flag warehouses running idle outside business hours
Scan for jobs configured to trigger outside their allowed window
Flag Lakeflow Declarative Pipelines with triggers set to fire at off-hours times
Define allowed hours for dev and test workspaces. Tag-based rules target the right environments.
Running resources are checked against the policy. Anything outside the window is flagged as a violation.
Out-of-window resources are stopped automatically. The resource creator and workspace admins are notified.
Connect a workspace and see what dev compute is running after hours.
Start Free Trial