Hi, production service down all afternoon since the GCP outage. Persistent SqliteError: attempt to write a readonly database (code: SQLITE_READONLY) crash loop. Tried everything from the dashboard: At first it said it was deployed but Post deploy never did anything.

Restart — same error

Redeployed older known-good commits — same error

Restored from a 9-hour-old backup (new volume created from backup) — same error on the brand-new volume

The fact that a freshly restored volume is also mounting read-only strongly suggests the underlying node is unhealthy. Per your FAQ, please move my volume to a healthy node or let me know if this is just because the system is down.

⚠️ STRICT SCOPE OF WORK — PLEASE READ CAREFULLY:

The ONLY action I'm requesting is: move the existing /data volume — with all its data intact — to a healthy node so it mounts read-write again.

DO NOT touch, modify, or change anything else. Specifically:

DATA & STORAGE:

DO NOT delete the volume

DO NOT delete, modify, or expire any backups (I have a daily snapshot from 9 hours ago and a manual snapshot from 19 hours ago — both must remain)

DO NOT reset, wipe, recreate, or reinitialize the database

DO NOT alter the contents of the /data volume in any way

CODE & DEPLOYMENT:

DO NOT modify, rebuild, or redeploy from any specific commit

DO NOT change the Dockerfile, build settings, or start command

DO NOT touch the GitHub integration or auto-deploy settings

DO NOT modify any source code (it's in my GitHub repo — I assume you cannot anyway, just confirming)

SERVICE CONFIGURATION:

DO NOT modify environment variables

DO NOT change the domain, custom domain settings, or DNS routing

DO NOT change networking, ports, or replicas

DO NOT modify the service, project, or environment settings

DO NOT delete or rebuild the service itself

If something is unclear, please reply in this thread and ask me before taking any action. I'd rather wait longer for a careful fix than have anything else broken.

Context: I run a real estate brokerage on this platform (NexPath). Every record — agents, transactions, revshare ledgers, CDAs, broker reports, documents — is critical business data. The system is otherwise working perfectly; this is purely a read-only volume mount issue it looks like from the GCP outage.

IDs:

Project ID: 42de96c2-5bf6-425c-bc70-6c5135f7c340

Service ID: d0f2734e-239b-47bf-9123-9237dbb387d3

Environment ID: d6c560ca-24c9-4a35-866d-6b5cf32e7be0

Service name: web

Domain: nexpathplatform.com / www.nexpathplatform.com

Mount path: /data

Thanks for your help.