Contents / Incidents & stability
Reading the incident records
Incidents lists every incident event newest-first: site, start time, recovery time, duration, error type and summary.
#Error types & where to look first
| Type | Meaning | Check first |
|---|---|---|
connect | TCP connection failed | Server down, firewall/security group, ports 80/443 |
timeout | Connect or response timed out | Server load, PHP-FPM/database congestion, attack traffic |
ssl | TLS handshake failed | Certificate chain, hostname match, TLS version negotiation |
ssl_expired | Certificate expired | Renew and reload immediately |
http_code | Abnormal status code | 5xx → check backend; 4xx → check config/paths |
keyword | Page content abnormal | White screen, DB-failure page, tampering |
page_error | Page prints a server error | The raw error usually names the file and line directly |
slow | Response too slow | Slow queries, CPU/IO |
node_fault | Your fetch channel is unavailable | Not counted as an incident; test the node under Nodes & proxies |
blocked_target | Resolved to an internal/reserved address | Check whether the domain's DNS was changed |
Tip
The troubleshooting hint under each error type is the platform's first-check list per type — read it before digging into server logs.
Related: Stability ranking · "It opens for me but got marked down"