Usage
Usage shows you how hard you’re working Saiku Cloud — request volume per hour, P50/P95/P99 query latency, error rate, file storage trending against your quota. The whole page is a small set of charts and a filter bar: pick a time range, hour or day bucketing, and the charts redraw.
It’s the right page to look at when you want to know whether you’re approaching rate-limit ceilings, whether last night’s spike in agent errors was a tenant-wide problem or one bad integration, or whether your storage is creeping toward your plan’s quota.
What you’ll see
Request volume. Bars of API calls per bucket. The bucket is hour by default for ranges of 7 days or less, day for longer.
Latency. Three lines for P50, P95, and P99 query latency. Use the P95 line for SLA discussions; P50 for typical user experience; P99 for the worst-case-pathological-query story.
Error rate. A percentage line showing what fraction of requests returned non-2xx. Healthy steady state is well under 1%; sustained spikes deserve a look in the audit log to see which paths and which actors are involved.
Storage quota. A simple bar showing your current file-upload usage vs. your plan’s limit. Same number as you see on the Files page header, surfaced here so it’s visible alongside everything else.
Filtering
The bar at the top lets you pick a time range (last hour, last 24h, last 7d, last 30d, or custom) and a bucket (hour or day). The range syncs to the URL so you can bookmark or share a specific view.
What you won’t see (yet)
- Per-cube breakdown. Tenant-wide rollup only in v1. The audit log has per-event detail if you need to slice by cube.
- Per-user breakdown. Same — audit log is the path to per-actor visibility.
- Real-time. The page reads from our metrics pipeline which lags the actual requests by about 30 seconds. Refresh to pull fresh numbers.