Incidents — LADLE status

Service incidents (vs security incidents) are logged here. Every one has a duration, an impact statement, and a resolution note. If it happened, it's here.

How to read this

Full outage = every user couldn't use LADLE. Partial degradation = some subset of users affected, or one feature down while the rest worked. Elevated latency = the app worked but was slow. Post-mortems are published for any incident lasting more than 30 minutes.

2026-07-14 — Partial degradation (browser-specific) (42 min)

Impact: Streaming responses stalled for approximately 1% of chats on Safari 17.5 due to an SSE-buffering interaction with Safari's HTTP/2 stack.

Resolution: Deployed a header-tuning change that keeps buffers flushing under Safari's threshold. Verified fix across Safari 17.5 and 17.6.

Date2026-07-14

Duration42 min

SeverityPartial degradation (browser-specific)

2026-05-02 — Full outage (8 min)

Impact: Full outage — application returned 503 across all endpoints. Root cause: our Vercel deployment coincided with a Vercel edge network incident. All requests failed until Vercel resolved on their end.

Resolution: No LADLE-side change required; Vercel resolved. Follow-up: added Vercel status monitoring to our on-call alerting so we're notified earlier next time.

Date2026-05-02

Duration8 min

SeverityFull outage

Uptime

See /status/uptime-history for the 30d / 90d / 365d rolling uptime figures. Target SLA is 99.5%; actual has been higher since launch.

REPORT AN ISSUE · HELLO@LADLE.CHAT

Every service incident.

How to read this

2026-07-14 — Partial degradation (browser-specific) (42 min)

2026-05-02 — Full outage (8 min)

Uptime

Silent incidents make repeat incidents.

Every service incident.

How to read this

2026-07-14 — Partial degradation (browser-specific) (42 min)

2026-05-02 — Full outage (8 min)

Uptime

Silent incidents make repeat incidents.