← Dashboard

System health

Snapshot freshness per backend. Cron/poll-run tables below are MD-side write audits — same data drives all three backends, so the cron stats are backend-agnostic.

MotherDuck (retired)
could not connect to database: Invalid Input Error: Invalid Error: Request failed: Your request is not authenticated. Pl
HistoricalACTIVE
Vercel Blob: This store does not exist.

Per-endpoint health over the last 24 hours: did the call succeed (HTTP 2xx, parseable JSON) and did the response contain the fields our parsers depend on? Schema drift catches an upstream API quietly dropping a field; success rate catches it failing entirely.

Outbound calls (24h)
0
0 failed
Success rate
100.00%
weighted across all endpoints
Schema drift
0
all clear
Most recent drift
—

Endpoint health

EndpointCallsSuccessAvg / MaxSchemaNotes
ea/list———ok—
ea/snapshot———ok—
ea/detail———ok—
cp/station_list———ok—
cp/station_info———ok—
blink/detail———ok—
blink/afdc———ok—
blink/pins———ok—
evgo/auth———ok—
evgo/sites———ok—
evgo/chargers———ok—
evgo/pricing———ok—
shell/nearby———ok—
evergo/getsiteinfo———ok—
evergo/chagersinview———ok—
google/places-text-search———ok—
google/places-details———ok—
google/places-autocomplete———ok—
google/routes-matrix———ok—
google/routes-compute———ok—

User-triggered routes (Google Places autocomplete/details, Routes API) appear here only when their containers happen to flush via a co-located write — most of their calls log first-occurrence to Vercel function logs instead.