Exercise 2: 30 Instances Firing the Same Alert at Once — Possible Solution
====================================================================

Using the chapter's own route config:

route:
  receiver: 'default-slack'
  group_by: ['alertname', 'job']
  group_wait: 30s
  group_interval: 5m
  repeat_interval: 4h

-- What happens when 30 instances fire the same alert within seconds --

Because group_by is set to ['alertname', 'job'], all 30 firing alerts
share the exact same alertname (say, HighErrorRate) and the exact
same job label (they're all instances of the same service) -- so
Alertmanager treats them as ONE group, not 30 separate incidents.
When the first of the 30 alerts arrives, Alertmanager starts the
group_wait timer (30s) rather than notifying immediately -- giving
the other 29 alerts, which are almost certainly about to arrive
within the same few seconds since they're all instances of the same
struggling service, time to be folded into that same group. Once
group_wait elapses, ONE notification is sent, listing all alerts
currently in the group -- not 30 separate pages or Slack messages.
If more instances of the same alert/job combination fire afterward,
group_interval (5m) governs how long Alertmanager waits before
sending an updated notification about the newly joined alerts, again
as one bundled update rather than one message per new alert.

-- Why this matters for alert fatigue --
--
-- Without grouping, 30 nearly-simultaneous alerts would produce 30
-- separate notifications for what is, in reality, a SINGLE incident
-- (the same underlying problem affecting every instance of one
-- service). Receiving 30 pages for one incident is exactly the kind
-- of noise that trains people to start ignoring or muting alerts
-- reflexively, the chapter's own stated definition of alert fatigue.
-- Grouping turns that single incident back into a single
-- notification, which is both less overwhelming and more accurate --
-- it correctly represents "one thing is wrong, affecting many
-- instances" rather than falsely implying 30 independent problems.

WHY THIS WORKS AS AN ANSWER
------------------------------
This walks through exactly how group_by/group_wait/group_interval
combine to turn 30 near-simultaneous alerts into one bundled
notification, then explicitly ties that behavior back to the
chapter's own definition of alert fatigue rather than just restating
that "grouping helps."