Troubleshooting

Diagnostic order

1. 1

   Record the exact message, URL, time, and request ID.

2. 2

   Check workspace, role, and subscription.

3. 3

   Check freshness and import state.

4. 4

   Review system status and known issues.

5. 5

   Retry once only after addressing the cause.

Do not hide the cause

• Do not repeat a write after an unknown outcome.
• Do not rotate a token before checking permissions and quotas.
• Do not clear browser data before preserving diagnostics.

Escalate

Open a ticket with redacted evidence. For full service unavailability include network, region, and a test from another connection.