Satirical GitHub contribution graph where red squares represent platform outage days, sourced live from mrshu/github-statuses via githubstatus.com.
Key Takeaways
GitHub had 35.1 days of downtime in the last year across 170 days with at least one incident.
Worst single day: Thu, Nov 20, 2025, with 1.1 days of cumulative outage logged.
Scheduled maintenance is excluded; data is fetched live and aggregated from reconstructed incident history.
Heatmap rendering uses Mantine; darker red squares indicate longer outage duration on a given day.
Hacker News Comment Review
The official githubstatus.com page and the third-party mrshu/github-statuses tracker show dramatically different numbers, raising questions about how GitHub scopes its SLA coverage by component.
Commenters flagged a methodology gap: the site may sum per-component downtime rather than deduplicate overlapping incidents, inflating day-level totals like the Nov 20 figure.
There is open debate about whether GitHub runs on Azure infrastructure at all, meaning the graph may say more about GitHub’s own ops than Azure’s reliability broadly.
Notable Comments
@natty: Highlights the legal tension between GitHub’s SLA terms and real-world uptime as tracked by third-party sources.
@hxtk: Explains the deduplication logic in mrshu’s tracker and why day totals can exceed 24 hours when component outages overlap.
