HLS su edge e monitoraggio ops con latenza p95.

Evita di saturare Puma in diretta, separa i check Rails/pubblico e traccia la latenza /up per alert più affidabili.

Co-authored-by: Cursor <cursoragent@cursor.com>

backend/spec/services/ops/health_checks_spec.rb

@@ -24,7 +24,7 @@ RSpec.describe Ops::HealthChecks do
       )
       %i[
         check_recordings_size check_postgres check_redis check_mediamtx check_garage
-        check_sidekiq_heartbeat check_sidekiq_dead check_http_public
+        check_sidekiq_heartbeat check_sidekiq_dead check_http_rails check_rails_latency
       ].each do |method|
         allow_any_instance_of(described_class).to receive(method).and_return(
           described_class::Finding.new(
@@ -33,6 +33,7 @@ RSpec.describe Ops::HealthChecks do
           )
         )
       end
+      allow_any_instance_of(described_class).to receive(:public_check_due?).and_return(false)
 
       described_class.new.call
 
@@ -50,7 +51,7 @@ RSpec.describe Ops::HealthChecks do
       )
       %i[
         check_recordings_size check_postgres check_redis check_mediamtx check_garage
-        check_sidekiq_heartbeat check_sidekiq_dead check_http_public
+        check_sidekiq_heartbeat check_sidekiq_dead check_http_rails check_rails_latency
       ].each do |method|
         allow_any_instance_of(described_class).to receive(method).and_return(
           described_class::Finding.new(
@@ -59,11 +60,40 @@ RSpec.describe Ops::HealthChecks do
           )
         )
       end
+      allow_any_instance_of(described_class).to receive(:public_check_due?).and_return(false)
 
       summary = described_class.new.summary
 
       expect(summary[:status]).to eq("ok")
-      expect(summary[:checks].size).to eq(9)
+      expect(summary[:checks].size).to eq(10)
+    end
+
+    it "degraded quando la latenza p95 supera la soglia warning" do
+      allow_any_instance_of(described_class).to receive(:build_findings).and_return([
+        described_class::Finding.new(
+          kind: "rails_latency", severity: "warning", healthy: false,
+          title: "Latenza elevata", message: "p95 2500ms", metadata: {}, fingerprint: "rails_latency:p95"
+        )
+      ])
+
+      expect(described_class.new.summary[:status]).to eq("degraded")
+    end
+  end
+
+  describe "#check_http_public" do
+    it "è warning, non critical" do
+      allow(Rails.env).to receive(:test?).and_return(false)
+      allow(MatchLiveTv).to receive(:app_public_url).and_return("https://www.matchlivetv.it")
+
+      checks = described_class.new
+      response = instance_double(Faraday::Response, status: 200)
+      allow(Faraday).to receive(:get).and_return(response)
+
+      finding = checks.send(:check_http_public)
+
+      expect(finding.severity).to eq("warning")
+      expect(finding.kind).to eq("http_public")
+      expect(finding.healthy).to be(true)
     end
   end
 end

backend/spec/services/ops/up_latency_tracker_spec.rb

@@ -0,0 +1,19 @@
+# frozen_string_literal: true
+
+require "rails_helper"
+
+RSpec.describe Ops::UpLatencyTracker do
+  before do
+    described_class.redis&.del(described_class::REDIS_KEY)
+  end
+
+  it "calcola il p95 sui campioni registrati" do
+    (1..20).each { |ms| described_class.record(ms * 10) }
+
+    expect(described_class.p95).to eq(190)
+  end
+
+  it "restituisce nil senza campioni" do
+    expect(described_class.p95).to be_nil
+  end
+end