[IMP] maintenance_service_http_monitoring : speed up KO confirmation

The previous design required two passes of the 10-minute sweep before a
maintenance request was created, so an outage took at least ~10 extra
minutes (up to 20) to be reported.

Split the monitoring into two crons with distinct responsibilities:

- the 10-minute sweep only discovers outages, flags the KO streak via
  ``http_first_ko_at`` and auto-resolves recovered services;
- a new 2-minute confirmation cron re-checks only the flagged services
  and creates a request once the service has been continuously KO for
  ``HTTP_KO_CONFIRMATION_DELAY`` (lowered from 5 to 2 minutes).

This keeps the transient-outage filter while removing the extra sweep
latency. Services already under an open request are skipped by the
confirmation cron and remain handled by the sweep.

maintenance_service_http_monitoring/README.md

@@ -39,13 +39,19 @@ By default, maintenance mode lasts 4 hours. To change this:
 
 ## Cron Jobs
 
-Two scheduled actions are installed:
+Three scheduled actions are installed:
 
 1. **HTTP Service Monitoring: check all services**
-   - Runs every 15 minutes
-   - Checks HTTP status of all active service instances with URLs
+   - Runs every 10 minutes
+   - Discovery sweep: checks HTTP status of all active service instances with URLs
+     and flags the start of a KO streak
 
-2. **HTTP Service Monitoring: deactivate expired maintenance mode**
+2. **HTTP Service Monitoring: confirm KO services**
+   - Runs every 2 minutes
+   - Re-checks only the services currently flagged KO (without an open request) and
+     creates a maintenance request once the outage is confirmed
+
+3. **HTTP Service Monitoring: deactivate expired maintenance mode**
    - Runs every 15 minutes
    - Automatically disables maintenance mode when the end time is reached
 
@@ -98,14 +104,18 @@ On service instances, you can see:
 ## Automatic Maintenance Requests
 
 When a service fails HTTP checks:
+- The 10-minute discovery sweep flags the outage (``http_first_ko_at``) but does
+  **not** create a request yet
+- The 2-minute confirmation cron re-checks the flagged services only. A request is
+  created once the service has been continuously KO for at least 2 minutes
+  (``HTTP_KO_CONFIRMATION_DELAY``), so a short transient outage is not flagged
 - A corrective maintenance request is created per failing service, named
   ``[HTTP KO] {service_url}``
 - The request description includes the error detail: the HTTP status code,
   or a network error label (timeout / DNS / SSL) when no HTTP response was received
 - No duplicate is created as long as an open request already exists for that service
-- A **double-check** is performed before creating the request: the service is retested
-  after 2 seconds. A maintenance request is only created if the service fails **both**
-  checks, reducing noise from transient HTTP errors
+- The confirmation cron skips services that already have an open request; those are
+  handled by the discovery sweep
 
 When a service recovers (returns HTTP 200 after having an open request):
 - The open maintenance request is automatically moved to the first **done** stage