`504` Gateway Timeouts on Coverage Uploads (Tracking Issue)

[Slach]

look details in https://github.com/Slach/clickhouse-dump/actions/runs/18029504406/job/51303534182#step:6:167

 v0.6.17

📄 Using coverage file: coverage.out
  ·job_flag: run-18029504406
🚀 Posting coverage data to https://coveralls.io/api/v1/jobs
HTTP error:
---
Error: Gateway Timeout (504)

[pgvishnuram]

we are also observing the same for the past 6 hours

[ehuelsmann]

CloudFlare locations Dallas and Amsterdam both report 504s for me, consistently.

[eonarheim]

Same here in Minneapolis, 504s since last night 8:40pm central US

[afinetooth]

All, Thank you for reporting. The issue was resolved around 6:15am PDT.

We are actively monitoring, and will continue working on mitigations until we can put an end to these.

You can track status here:
https://status.coveralls.io/ (current incident)

Two things:

1. If you are experiencing put 504s—in particular 504 errors of coverage uploads, please report in this issue; and you can follow the same issue for updates. We'll do our best to circle back here and other issues.
2. The most helpful report details, right now, are a timestamp, or at least timeframe, on your error, and any Cloudflare Ray ID you may have received with it.

---

References:

• Tracking issue (>>THIS ISSUE<<): 504 Gateway Timeouts on Coverage Uploads (Tracking Issue) #1824
• Status Page: https://status.coveralls.io/
• Incident: https://status.coveralls.io/incidents/qrwpyt8h4vf1

[paulromano]

I'm still seeing 504s sporadically, e.g., here's one that I just got:
https://github.com/openmc-dev/openmc/actions/runs/18198604615/job/51811431543

[afinetooth]

Thanks for reporting, @paulromano.

We've applied mitigations over the past few days that have significantly reduced these 504s, but during our overnight (pre 6am PDT), where they are addresses by automations, we still aren't able to recover from every case due to the nature of the massively parallel requests. It takes an effective 15-20 min before the automation resolves the environment conditions on specific servers that are causing the timeouts.

We continue to work on this, so I'm hopeful "sporadic" will soon becomes "infrequent" -> "rare" -> "very rare" -> "never". That's the goal.

[lrivkin]

Experiencing elevated 504s today for many builds between Friday, October 3rd, 2025 at 20:15:05 UTC and now (Friday, October 3rd, 2025 at 21:50:31 UTC)