Google Cloud Outage June 2025 - Gmail, Spotify Down 7 Hours | Null Pointer Crash Explained
June 12, 2025: Google Cloud Platform experienced a 7+ hour catastrophic outage that took down Gmail, Google Docs, Google Drive, Maps, Spotify, Snapchat, Discord, and thousands of other services. The cause? A null pointer error—one of the most basic programming mistakes—in a policy configuration change. Even Google's world-class engineers can deploy fundamental coding errors to production. And when they do, billions of users experience the consequences.
☁️ THE OUTAGE AT A GLANCE:
7+ hours of widespread service disruption
Affected multiple regions: North America, Europe, Asia-Pacific, South America
Google Service Control crash loop prevented automatic recovery
Null pointer dereference in policy configuration change
Gmail, Docs, Drive, Maps, YouTube all down/degraded
Spotify offline - 600M+ users unable to stream music
Snapchat unreachable - hundreds of millions affected
Discord down AGAIN (uses Google Cloud infrastructure)
Even Cloudflare services disrupted (GCP dependencies)
Fast detection (2 minutes) but slow recovery (7+ hours)
⏱️ TIMELINE OF FAILURE:
June 12, 2025
~12:00 PM PT (3:00 PM ET) - Policy change deployed to Service Control
12:15 PM - Gmail errors begin, email sending/receiving fails
12:20 PM - Google Docs & Drive unreachable, documents won't load
12:25 PM - Google Maps fails, navigation stops mid-route
12:30 PM - YouTube buffering indefinitely, Spotify offline
12:30 PM - Status page updated: "Investigating GCP issues"
1:00 PM - Multiple regions confirmed affected
2:00 PM - Some regions begin partial recovery
5:00 PM+ - Gradual restoration continues
7+ hours - Full recovery for most services
Total: Over 7 hours of major disruption, with regional variation in recovery times.
⚙️ WHAT ACTUALLY BROKE:
*The Root Cause:*
Policy configuration change to Google Service Control → introduced null pointer dereference → Service Control crashes when accessing empty memory location
*The Crash Loop:*
Service Control crashes → automatically restarts → loads same bad configuration → crashes again → repeats infinitely
Normal failure: Crash, restart, recover (seconds of downtime)
Crash loop: Crash, restart, crash, restart, crash... (hours of downtime)
*Why It Spread Globally:*
Configuration deployed to Service Control across multiple regions simultaneously. Every region that received the update entered the crash loop.
Regional redundancy doesn't help when the failure is replicated to all regions.
*What Service Control Does:*
"Traffic cop" for Google Cloud API requests. Validates every API call, checks permissions, manages quotas. When Service Control fails, APIs can't validate requests → services can't function.
Dependency chain: All GCP services → Service Control → Null pointer crash → Everything fails
🌐 WHO WAS AFFECTED:
*Google's Own Services:*
Gmail - email sending/receiving failed for billions
Google Docs & Drive - documents inaccessible, collaboration stopped
Google Maps - navigation failed, directions unavailable
YouTube - videos wouldn't play, buffering indefinitely
Google Meet - video conferencing failed, virtual meetings cancelled
*Services Built on Google Cloud:*
Spotify - 600M+ users, music streaming completely offline
Snapchat - hundreds of millions unable to send snaps/view stories
Discord - down for second time in months (Episode 7 = Discord's own issues, this = GCP dependency)
PayPal (partial impact)
Twitter/X (some services affected)
Home Depot, Target (e-commerce disruptions)
*Infrastructure Services:*
Cloudflare - some services disrupted because they use GCP
Other CDN/security providers with GCP dependencies
*Result:* Billions of users affected globally. Businesses unable to operate. Productivity halted for 7+ hours.
💰 THE PRODUCTIVITY CATASTROPHE:
*Google Workspace Impact:*
Companies that migrated entirely to Google Workspace experienced complete productivity shutdown:
No email for 7 hours
No access to documents/spreadsheets/presentations
No video conferencing
No collaboration tools
For remote-first companies, this meant zero ability to coordinate work.
*The Dependency Problem:*
Unlike AWS/Azure where you manage infrastructure, Google Workspace = fully managed. You're completely dependent on Google's services working. No control, no mitigation, just wait.
THE BREAKDOWN ECONOMY - Season 2 Complete:
Ep 1: CrowdStrike - 8.5M Computers
Ep 2: AT&T - 100M Without 911
Ep 3: Change Healthcare - $872M Ransomware
Ep 4: Meta - 3B Users Locked Out
Ep 5: Ticketmaster - Monopoly Failure
Ep 6: Southwest - 1990s System Meltdown
Ep 7: Discord - Gaming Went Silent
Ep 8: AWS - 15-Hour Cloud Failure
Ep 9: Cloudflare - Broke the Internet
Ep 10: Google Cloud - 7-Hour Null Pointer (This Episode)
Season 2 Finale drops next week: What 10 major failures tell us about the breakdown economy.
🔔 SUBSCRIBE for infrastructure failure analysis
💬 COMMENT: Were you affected by the Google Cloud outage?