Skip to main content

sre.uptime_timeseries

Status snapshots for each monitored site, one row per upptime bot commit.

One row per upptime bot commit per site. The upptime bot commits to the cds-snc/status-statut history YAML files once per day (~23:2x UTC) as a daily snapshot, plus additional commits immediately when a status change is detected mid-day.

Provenance

Fully automated. No manual upload required. Sourced from the GitHub Commits API for each site’s history file (history/<slug>.yml).

  • Updated: Daily, at ~04:00 ET
  • Retention: 7 years
  • Steward: EDCP Data and Research Team
  • Contact: Yinlin Zhang (Data and Research team)

Expected row count per site per day:

Scenario Expected rows
Normal all-up day 1 (daily ~23:2x UTC snapshot, status = 'up')
Outage day 2+ (down event + recovery event + daily snapshot)

Fields

Field Name Type Description
date varchar Hive partition key: the UTC date of the commit (YYYY-MM-DD). Use in WHERE date = '...' or WHERE CAST(date AS DATE) BETWEEN ... filters.
ping_timestamp varchar UTC timestamp of the upptime bot’s commit (YYYY-MM-DD HH:MM:SS). On a normal day this is the daily snapshot at ~23:2x UTC. On an outage day earlier commits record when the site went down or recovered.
slug varchar Upptime site identifier, same six values as in sre.uptime_summary.
lang varchar Short label, same six values as in sre.uptime_summary.
status varchar Ping result at the moment of this commit: up, down, or unknown.
http_code bigint HTTP status code at the moment of this commit. 200 on a normal day; 0 or 503 when down. null if unparseable.
response_ms bigint Response time in milliseconds at the moment of this commit. Single-ping measurement at commit time, not a daily average. null if unparseable.

Notes

response_ms is a single-ping measurement. It records the response time at the exact moment of the upptime bot’s commit, not a daily average. For averaged response times use sre.uptime_summary (response_day_ms, response_week_ms, etc.), which are computed from all 288 five-minute pings.

status = 'down' rows count detection events, not outage duration. Each down row is one commit, one moment when upptime detected a failure. For total minutes of downtime on a given date, use the daily_minutes_down JSON column in sre.uptime_summary.

Upstream monitoring gap (July 28 – September 23, 2026). No automated ping commits were made to cds-snc/status-statut during this period because synthetic monitoring was halted. The table contains zero rows between July 28 and September 23, 2026.