2026-10-09 10:43 UTCIncidents Detected

Google Cloud Storage Recovers

Early on Friday, October 9, there is no widespread internet disruption. The main story from the past day is an outage at Google Cloud Storage that lasted about 11 hours and ended overnight.

**Google Cloud Storage (resolved).** From 12:19 UTC to 23:30 UTC on October 8, Google Cloud Storage in the us-central1 region returned elevated latency and errors on upload and write operations. A smaller share of reads was also slower or failed. Google said no workaround was available. Mitigations began showing improvement around 17:57 UTC, and the incident was closed late in the evening UTC. Anything that writes to buckets in us-central1, including backups, data pipelines and apps that store user uploads, may have seen failed or retried writes during that window. ([Google Cloud status](status.cloud.google.com))

**Other incidents in the past 72 hours:**

**GitHub** had a rough run of incidents. On October 6, several services were degraded for about 3h40m (19:57 to 23:40 UTC). On October 7 there were two critical incidents affecting Git operations, pull requests and Actions, from 15:14 to 16:25 UTC and from 17:17 to 18:04 UTC. All are resolved. ([GitHub status](githubstatus.com))

**Anthropic (Claude):** On October 7, a spend-limit bug wrongly paused some organizations and caused refused requests across the Claude API, claude.ai and Claude Code. That issue is resolved. Usage data on platform.claude.com is still in monitoring, and the latest ~30 minutes may show up late; API requests are not affected. ([Claude status](status.claude.com))

**Yandex Cloud:** A power supply problem hit the ru-central1-b availability zone on October 8, as covered in yesterday's report.

**MTN Nigeria:** A network disruption that started October 8 has been declared resolved. MTN advised customers who still have problems to restart their phones.

**Ongoing, low impact:** Akamai is still monitoring origin connectivity errors in APAC (open since September 22). Twilio lists several carrier-specific SMS/MMS delivery delays, the newest affecting China Mobile. Zoom is still working on Polycom devices going offline. Cloudflare shows routine rerouting at a number of individual data centers, with no customer-facing incident beyond two minor Cloudflare One/WARP issues.

**All clear:** AWS, Azure, Azure DevOps, GitHub (now), OpenAI, Discord, Atlassian, Slack, Vercel, Heroku, DigitalOcean, Stripe, Cursor, Groq, Replicate, ElevenLabs, Cohere and Stability AI report normal operation. We found no notable BGP or DNS events in public reporting.

Crowd-sourced trackers show scattered early-morning spikes for services such as Verizon, Wells Fargo and Microsoft Teams. None of them has a confirmed provider-side incident, so treat these as unverified.

**Bottom line:** The internet is up. Google Cloud Storage writes in us-central1 have recovered after a long outage on October 8.

← Home View all history

Google Cloud Storage Recovers

Last 30 Days