Aggiunge monitoraggio ops: health check, log scanner, dashboard e notifiche.

Introduce incidenti con dedup, job Sidekiq ogni 3 min, scan log cron, /up/deep,
dashboard /admin/ops e push ntfy; include Sentry opzionale e fix test suite.

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

backend/spec/services/ops/health_checks_spec.rb

@@ -0,0 +1,69 @@
+require "rails_helper"
+
+RSpec.describe Ops::HealthChecks do
+  describe "#call" do
+    it "risolve incidenti quando il disco è sano" do
+      Ops::Incident.create!(
+        kind: "disk_space",
+        severity: "warning",
+        status: "open",
+        title: "Disco",
+        message: "alto",
+        metadata: {},
+        fingerprint: "disk_space:root",
+        occurrence_count: 1,
+        first_seen_at: Time.current,
+        last_seen_at: Time.current
+      )
+
+      allow_any_instance_of(described_class).to receive(:check_disk_root).and_return(
+        described_class::Finding.new(
+          kind: "disk_space", severity: "info", healthy: true,
+          title: "OK", message: "OK", metadata: {}, fingerprint: "disk_space:root"
+        )
+      )
+      %i[
+        check_recordings_size check_postgres check_redis check_mediamtx check_garage
+        check_sidekiq_heartbeat check_sidekiq_dead check_http_public
+      ].each do |method|
+        allow_any_instance_of(described_class).to receive(method).and_return(
+          described_class::Finding.new(
+            kind: "test", severity: "info", healthy: true,
+            title: "OK", message: "OK", metadata: {}, fingerprint: "#{method}:ok"
+          )
+        )
+      end
+
+      described_class.new.call
+
+      expect(Ops::Incident.find_by(fingerprint: "disk_space:root").status).to eq("resolved")
+    end
+  end
+
+  describe "#summary" do
+    it "restituisce status ok quando tutti i check sono sani" do
+      allow_any_instance_of(described_class).to receive(:check_disk_root).and_return(
+        described_class::Finding.new(
+          kind: "disk_space", severity: "info", healthy: true,
+          title: "OK", message: "OK", metadata: {}, fingerprint: "disk_space:root"
+        )
+      )
+      %i[
+        check_recordings_size check_postgres check_redis check_mediamtx check_garage
+        check_sidekiq_heartbeat check_sidekiq_dead check_http_public
+      ].each do |method|
+        allow_any_instance_of(described_class).to receive(method).and_return(
+          described_class::Finding.new(
+            kind: "test", severity: "info", healthy: true,
+            title: "OK", message: "OK", metadata: {}, fingerprint: "#{method}:ok"
+          )
+        )
+      end
+
+      summary = described_class.new.summary
+
+      expect(summary[:status]).to eq("ok")
+      expect(summary[:checks].size).to eq(9)
+    end
+  end
+end

backend/spec/services/ops/incident_recorder_spec.rb

@@ -0,0 +1,55 @@
+require "rails_helper"
+
+RSpec.describe Ops::IncidentRecorder do
+  describe ".record" do
+    it "crea un incidente nuovo" do
+      incident = described_class.record(
+        kind: "disk_space",
+        severity: "critical",
+        title: "Disco pieno",
+        message: "90% usato",
+        fingerprint: "disk_space:root"
+      )
+
+      expect(incident).to be_persisted
+      expect(incident.occurrence_count).to eq(1)
+      expect(incident.status).to eq("open")
+    end
+
+    it "deduplica per fingerprint incrementando occorrenze" do
+      described_class.record(
+        kind: "disk_space",
+        severity: "critical",
+        title: "Disco pieno",
+        message: "90%",
+        fingerprint: "disk_space:root"
+      )
+      second = described_class.record(
+        kind: "disk_space",
+        severity: "critical",
+        title: "Disco pieno",
+        message: "91%",
+        fingerprint: "disk_space:root"
+      )
+
+      expect(Ops::Incident.where(fingerprint: "disk_space:root").count).to eq(1)
+      expect(second.occurrence_count).to eq(2)
+    end
+  end
+
+  describe ".resolve" do
+    it "chiude incidenti aperti con la stessa fingerprint" do
+      incident = described_class.record(
+        kind: "disk_space",
+        severity: "warning",
+        title: "Test",
+        message: "msg",
+        fingerprint: "disk_space:root"
+      )
+
+      described_class.resolve(fingerprint: "disk_space:root")
+
+      expect(incident.reload.status).to eq("resolved")
+    end
+  end
+end

backend/spec/services/ops/log_pattern_matcher_spec.rb

@@ -0,0 +1,24 @@
+require "rails_helper"
+
+RSpec.describe Ops::LogPatternMatcher do
+  subject(:matcher) do
+    described_class.new(
+      patterns: [
+        { name: "http_500", severity: "warning", regex: /Completed 500/i },
+        { name: "disk_full", severity: "critical", regex: /No space left on device/i }
+      ]
+    )
+  end
+
+  it "trova pattern nei log" do
+    log = <<~LOG
+      I, [2026-06-10T19:45:10] Completed 500 Internal Server Error
+      E, [2026-06-10T19:45:11] No space left on device - /recordings
+    LOG
+
+    matches = matcher.scan(log)
+
+    expect(matches.map(&:name)).to contain_exactly("http_500", "disk_full")
+    expect(matches.find { |m| m.name == "disk_full" }.severity).to eq("critical")
+  end
+end

backend/spec/services/ops/log_scanner_spec.rb

@@ -0,0 +1,9 @@
+require "rails_helper"
+
+RSpec.describe Ops::LogScanner do
+  it "crea incidenti dai pattern trovati" do
+    log = Rails.root.join("spec/fixtures/ops/sample.log").read
+
+    expect { described_class.new.ingest(log) }.to change(Ops::Incident, :count).by_at_least(2)
+  end
+end

backend/spec/services/ops/notifier_spec.rb

@@ -0,0 +1,43 @@
+require "rails_helper"
+
+RSpec.describe Ops::Notifier do
+  let(:incident) do
+    Ops::Incident.create!(
+      kind: "test",
+      severity: "critical",
+      status: "open",
+      title: "Test alert",
+      message: "Messaggio di prova",
+      metadata: {},
+      fingerprint: "test:#{SecureRandom.hex(4)}",
+      occurrence_count: 1,
+      first_seen_at: Time.current,
+      last_seen_at: Time.current
+    )
+  end
+
+  describe "#notify" do
+    it "invia POST a ntfy quando configurato" do
+      conn = instance_double(Faraday::Connection)
+      response = instance_double(Faraday::Response, success?: true, status: 200, body: "")
+      allow(Faraday).to receive(:new).and_return(conn)
+      allow(conn).to receive(:post).and_return(response)
+      allow(ENV).to receive(:[]).and_call_original
+      allow(ENV).to receive(:[]).with("OPS_NTFY_URL").and_return("https://ntfy.example/ops")
+
+      described_class.new.notify(incident)
+
+      expect(conn).to have_received(:post).with("https://ntfy.example/ops", anything, hash_including("Title"))
+    end
+  end
+
+  describe ".notify_severity?" do
+    it "rispetta OPS_NOTIFY_SEVERITIES" do
+      allow(ENV).to receive(:fetch).and_call_original
+      allow(ENV).to receive(:fetch).with("OPS_NOTIFY_SEVERITIES", "critical").and_return("critical,warning")
+
+      expect(described_class.notify_severity?("critical")).to be(true)
+      expect(described_class.notify_severity?("info")).to be(false)
+    end
+  end
+end