cortazar.labs / status · 2026-06-03 17:25:24Z
All systems operational.
The collectors, the bus, the lakehouse, and the signals layer. Refreshed every 60s; this page is read off the same Trino tables the desk reads off.
last incident
3h ago · minor
collectors · 5 venues
Real-time ingest health.
ingest lag · last 24h
Tape latency, end-to-end.
From exchange WebSocket to bronze write, p50 across all collectors. The small bump near the right edge is the bitget reconnect.
uptime · last 90 days
One square per day.
Aggregated across all five collectors and core components. Hover for the date.
operational degraded incident
core components
The plant, piece by piece.
redpanda · bus
3 topics · 1 broker · 4.1 MiB/s
flink · streaming
lag ~ 380 ms · 0 failed checkpoints / 24h
iceberg · bronze
5.5B rows / 7d · last compaction 03:18Z
trino · query
p99 14 ms · 2 workers
k3s · rack
3 nodes · 0 alerts
openbao · secrets
rotated 2026-05-12
cloudflare · edge
ams-01 · 4 origins
signals · publisher
6 live · 1 in review (cvd-fast)
recent events · last 90 days
Incidents, releases, and planned work.
2026-05-19 11:38Z
venue · bitgetminor
WebSocket reconnect after 5m drop. Auto-recovered; no data loss; gap backfilled from REST snapshot.
2026-05-12 14:38Z
venue · bitgetminor
WebSocket reconnect after 3m drop. Auto-recovered.
2026-05-08 03:18Z
iceberg · bronzeplanned
Scheduled compaction completed. 18 minutes; reads served from prior snapshot.
2026-05-04 09:00Z
signalsrelease
Spoof score v6 promoted to live after 60d out-of-sample validation.
2026-04-30 22:14Z
k3s · node-02planned
Kernel update; workloads drained. 6m partial degradation on long-running queries.
2026-04-18 11:02Z
signalsrelease
Toxicity definition revised — refresh-driven cancels excluded from numerator.