fix(health): unconfigured Slack is UNKNOWN, never DOWN #13

2026-10-15T04:54:41Z

KofTwentyTwo commented

2026-10-15 04:54:41 +00:00

(Migrated from github.com)

This fixes a production finding: the slack health indicator reported an unconfigured Slack app (no SLACK_* secrets) as DOWN. That failed kof22's readiness probe and blocked a deploy; WB#250 added a temporary filter.

Fix

No Slack app secrets at all: the app readiness check is now NOT_CONFIGURED. Health reports UNKNOWN, which returns HTTP 200, with configured: false and the message "Slack is not configured". This state still appears in the readiness and central-health views.
Readiness not yet run (before the first scheduled check): health is UNKNOWN. It used to be DEGRADED.
Configured but broken (some secrets missing, or an invalid token key): health stays DOWN.
Configured and working: health is unchanged, UP or DEGRADED.
Docs: DESIGN section 8 is updated.

The core health module (RC.7) has no OUT_OF_SERVICE status. Its values are UP, DOWN, DEGRADED and UNKNOWN, and only DOWN maps to a 503, so UNKNOWN is the right status here.

Tests

not configured gives UNKNOWN with configured=false;
never checked gives UNKNOWN;
partially configured gives DOWN;
the existing invalid-key test still gives DOWN.

verify is green locally with 178 tests.

🤖 Generated with Claude Code

This fixes a production finding: the `slack` health indicator reported an unconfigured Slack app (no `SLACK_*` secrets) as DOWN. That failed kof22's readiness probe and blocked a deploy; WB#250 added a temporary filter. ## Fix - **No Slack app secrets at all:** the `app` readiness check is now `NOT_CONFIGURED`. Health reports **UNKNOWN**, which returns HTTP 200, with `configured: false` and the message "Slack is not configured". This state still appears in the readiness and central-health views. - **Readiness not yet run** (before the first scheduled check): health is **UNKNOWN**. It used to be DEGRADED. - **Configured but broken** (some secrets missing, or an invalid token key): health stays **DOWN**. - **Configured and working:** health is unchanged, UP or DEGRADED. - **Docs:** DESIGN section 8 is updated. The core health module (RC.7) has no OUT_OF_SERVICE status. Its values are UP, DOWN, DEGRADED and UNKNOWN, and only DOWN maps to a 503, so UNKNOWN is the right status here. ## Tests - not configured gives UNKNOWN with `configured=false`; - never checked gives UNKNOWN; - partially configured gives DOWN; - the existing invalid-key test still gives DOWN. `verify` is green locally with 178 tests. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: QRun-IO/qbit-slack#13