правки в транспорт сообщений + немного мониторинга

backlog.md

@@ -0,0 +1,9 @@
+В разработку:
+1. Создание нового диалога через UI
+2. При создании пользователя номер телефона копировать в профиль Russian_Phone
+3. Убрать хеширование устройства клиента в devise_json - надо видеть его параметры.
+4. Добавить параметр, ограничивающих кол-во неуспешных попыток ввода смс. 
+5. Поправить чтение сообщений от Битрикса. Сейчас они выглядят так: "[b]Антон Пичугин:[/b] [br]опять ты?" Надо убрать из текста сообщения Отправителя в битриксе
+
+Выполнено:
+1. Изменение в БД по аудитам (заполнение IP, сквозное заполнение UserSession)

codebase/backend/bitrix-local-app/app/main.py

@@ -159,7 +159,7 @@ def retry_delay(attempt: int, maximum: int) -> float:
 
 
 def safely_retryable(exc: Exception) -> bool:
-    return isinstance(exc, (httpx.ConnectError, httpx.TimeoutException)) or (
+    return isinstance(exc, (TimeoutError, httpx.ConnectError, httpx.TimeoutException)) or (
         isinstance(exc, httpx.HTTPStatusError) and exc.response.status_code in {429, 502, 503, 504}
     )
 
@@ -751,7 +751,10 @@ def create_app(settings: Settings | None = None) -> FastAPI:
                         ),
                     )
         try:
-            result = await deliver_outbound(request.app, row.id)
+            result = await asyncio.wait_for(
+                deliver_outbound(request.app, row.id),
+                timeout=cfg.bitrix_http_timeout_sec,
+            )
         except Exception as exc:
             logger.exception(
                 "outbound delivery failed",
@@ -1257,7 +1260,10 @@ async def process_outbound(app: FastAPI) -> None:
     if not row:
         return
     try:
-        await deliver_outbound(app, row.id)
+        await asyncio.wait_for(
+            deliver_outbound(app, row.id),
+            timeout=app.state.settings.bitrix_http_timeout_sec,
+        )
     except Exception as exc:
         logger.exception(
             "outbound retry failed",

codebase/backend/bitrix-local-app/tests/test_core.py

@@ -208,6 +208,7 @@ def test_fingerprint_ignores_signed_query_and_portal_validation():
 
 
 def test_network_timeouts_are_retryable():
+    assert safely_retryable(TimeoutError("Delivery operation timed out"))
     assert safely_retryable(httpx.ReadTimeout("Bitrix response timed out"))
     assert safely_retryable(httpx.ConnectTimeout("Bitrix connection timed out"))
 

deploy-steps.md

@@ -8,7 +8,7 @@
         - (без спец.символов) openssl rand -hex 16
         - openssl rand -base64 16 | xclip -selection clipboard # Linux
 
-Туннель до БД: ssh -i C:\Users\MI\.ssh\hansel -L 5433:192.168.0.211:5433 root@135.106.164.58 -N
+Туннель до БД: ssh -i C:\Users\MI\.ssh\hansel -L 5433:192.168.0.211:5432 root@135.106.164.58 -N
 
 #Обновление проекта
 
@@ -90,6 +90,28 @@ mkdir -p ~/.postgresql/
 wget https://storage.dbaas.selcloud.ru/CA.pem -O ~/.postgresql/root.crt
 chmod 0600 ~/.postgresql/root.crt
 
+### Правки DNS-маршрутизации (в прошлый раз, какие-то трабблы были с базовой маршрутизацией) - скрипт ниже создаст 
+sudo tee /etc/netplan/99-han-dns.yaml <<'EOF'
+network:
+  version: 2
+  ethernets:
+    eth0:
+      nameservers:
+        addresses:
+          - 8.8.8.8
+          - 1.1.1.1
+        search:
+          - selcloud.ru
+    eth1:
+      dhcp4-overrides:
+        use-dns: false
+        use-domains: false
+EOF
+
+sudo chmod 600 /etc/netplan/99-han-dns.yaml
+sudo netplan apply
+resolvectl flush-caches
+
 ### Проверка подключения к Postgers в ВМ
 nc -zv 192.168.0.211 5433
 
@@ -292,3 +314,132 @@ docker compose --env-file .env ps
 docker compose --env-file .env logs --tail=200 <SERVICE_NAME>
 docker inspect "$(docker compose --env-file .env ps -q <SERVICE_NAME>)"
 ```
+
+## Первоначальный выпуск TLS-сертификата
+
+Для ACME требуется работающий nginx по HTTP. В `.env` оставьте
+`NGINX_TLS_ENABLED=true`, но первый nginx запустите с временным переопределением:
+
+```sh
+NGINX_TLS_ENABLED=false \
+  docker compose --env-file .env up -d frontend-static nginx
+```
+
+Проверьте HTTP:
+
+```sh
+curl -I http://chat.han0107.ru/
+```
+
+Сначала рекомендуется проверить Certbot через staging:
+
+```sh
+docker compose --env-file .env --profile certbot run --rm certbot certonly \
+  --staging \
+  --webroot -w /var/www/certbot \
+  -d chat.han0107.ru \
+  --cert-name chat.han0107.ru-staging \
+  --email ap@han.ru \
+  --agree-tos --no-eff-email --non-interactive
+```
+
+После успешного staging-теста выпустите рабочий сертификат с основным cert-name
+без `--staging`:
+
+```sh
+docker compose --env-file .env --profile certbot run --rm certbot certonly \
+  --webroot -w /var/www/certbot \
+  -d chat.han0107.ru \
+  --cert-name chat.han0107.ru \
+  --email ap@han.ru \
+  --agree-tos --no-eff-email --non-interactive
+```
+
+Пересоздайте nginx уже с TLS:
+
+```sh
+docker compose --env-file .env up -d --force-recreate nginx
+docker compose --env-file .env exec -T nginx nginx -t -c /tmp/nginx.conf
+curl -I https://chat.han0107.ru/
+```
+
+Повторно запустите VM setup, чтобы он обнаружил проект и установил systemd-таймер
+продления сертификата:
+
+```sh
+sudo /opt/han-chat/backend/deployment/scripts/setup-vm.sh
+systemctl status han-chat-ssl-renew.timer
+
+## Запуск всего контура
+
+Теперь можно привести весь проект к состоянию, описанному Compose:
+
+```sh
+cd /opt/han-chat/backend
+docker compose --env-file .env up -d
+docker compose --env-file .env ps
+```
+
+Проверьте, что контейнеры не перезапускаются:
+
+```sh
+docker compose --env-file .env ps
+docker compose --env-file .env logs --since=10m
+```
+
+
+## Публичная проверка
+
+Запустите smoke-тест:
+
+```sh
+cd /opt/han-chat/backend
+deployment/scripts/smoke.sh
+```
+
+Также вручную проверьте:
+
+```sh
+curl -fsS https://chat.han0107.ru/api/v1/public/app-config | jq
+curl -fsS https://chat.han0107.ru/api/v1/public/content | jq
+curl -fsS \
+  https://chat.han0107.ru/auth/realms/han-chat/.well-known/openid-configuration | jq
+```
+
+Внутренний API не должен быть опубликован:
+
+```sh
+curl -i https://chat.han0107.ru/internal/safety/v1/messages/check
+```
+
+Ожидаемый статус — `404`.
+
+Откройте в браузере:
+
+```text
+https://chat.example.ru/
+```
+
+Для тестовой авторизации используйте значение `KEYCLOAK_OTP_MOCK_CODE` из `.env`.
+
+
+Настройки keykcloack:
+
+
+
+
+Для битрикса код установки приложения:
+docker compose --env-file .env exec -T api-backend python - <<'PY'
+import os
+import urllib.request
+
+base = os.environ["BITRIX_LOCAL_APP_BASE_URL"]
+token = os.environ["BITRIX_LOCAL_APP_INTERNAL_TOKEN"]
+
+request = urllib.request.Request(
+    base + "/internal/openlines/v1/setup/retry",
+    method="POST",
+    headers={"Authorization": "Bearer " + token},
+)
+print(urllib.request.urlopen(request).read().decode())
+PY

faq.md

@@ -0,0 +1,28 @@
+# Логи по контейнерам
+docker compose --env-file .env logs --tail 100 keycloak
+docker compose --env-file .env logs --tail 100 api-backend
+docker compose --env-file .env logs --tail 100 keycloak
+
+
+
+# Пересобрать образы и поднять всё заново
+Из каталога проекта на ВМ:
+cd /opt/han-chat/backend
+1. Собрать все локальные образы (как в DEPLOYMENT_GUIDE)
+docker compose --env-file .env build --pull
+2. Поднять весь контур
+docker compose --env-file .env up -d
+Или одной командой:
+cd /opt/han-chat/backend
+docker compose --env-file .env up -d --build
+
+# Только пересоздать контейнеры (без пересборки образов)
+Если меняли только .env (например, IP БД), образы пересобирать не нужно:
+cd /opt/han-chat/backend
+docker compose --env-file .env up -d --force-recreate
+
+# Полная пересборка + пересоздание
+cd /opt/han-chat/backend
+docker compose --env-file .env up -d --build --force-recreate
+Проверка
+docker compose --env-file .env ps

ops-monitoring/diagnose-han-chat.sh

@@ -0,0 +1,64 @@
+#!/usr/bin/env bash
+# diagnose-han-chat.sh — запускать из /opt/han-chat/backend
+set -euo pipefail
+cd /opt/han-chat/backend
+
+echo "=== 1. Статус сервисов ==="
+docker compose --env-file .env ps api-backend keycloak redis message-safety
+
+echo
+echo "=== 2. Healthcheck api-backend (что именно падает) ==="
+docker compose --env-file .env exec -T api-backend \
+  python -c "
+import json, urllib.request
+try:
+    r = urllib.request.urlopen('http://127.0.0.1:8000/health/ready', timeout=5)
+    print('HTTP', r.status)
+    print(json.dumps(json.loads(r.read()), indent=2, ensure_ascii=False))
+except urllib.error.HTTPError as e:
+    print('HTTP', e.code)
+    print(json.dumps(json.loads(e.read()), indent=2, ensure_ascii=False))
+"
+
+echo
+echo "=== 3. Keycloak health (management port 9000) ==="
+docker compose --env-file .env exec -T keycloak bash -c \
+  "exec 3<>/dev/tcp/127.0.0.1/9000 && printf 'GET /auth/health/ready HTTP/1.0\r\n\r\n' >&3 && cat <&3" \
+  2>/dev/null | head -20 || echo "Keycloak health FAILED"
+
+echo
+echo "=== 4. api-backend -> Keycloak (JWKS) ==="
+docker compose --env-file .env exec -T api-backend \
+  python -c "
+import os, urllib.request, json
+base = os.environ.get('KEYCLOAK_INTERNAL_URL','http://keycloak:8080/auth').rstrip('/')
+realm = os.environ.get('KEYCLOAK_REALM','han-chat')
+url = f'{base}/realms/{realm}/.well-known/openid-configuration'
+try:
+    d = json.loads(urllib.request.urlopen(url, timeout=5).read())
+    print('discovery OK, jwks_uri:', d.get('jwks_uri'))
+except Exception as e:
+    print('discovery FAILED:', e)
+"
+
+echo
+echo "=== 5. Счётчик рестартов ==="
+docker inspect --format '{{.Name}} restarts={{.RestartCount}} started={{.State.StartedAt}} health={{if .State.Health}}{{.State.Health.Status}}{{else}}n/a{{end}}' \
+  han-chat-api-backend-1 han-chat-keycloak-1
+
+echo
+echo "=== 6. Последние ошибки в логах ==="
+echo "--- api-backend ---"
+docker logs --tail 80 han-chat-api-backend-1 2>&1 | grep -iE 'error|failed|exception|jwks|settings|postgres' || echo "(нет совпадений)"
+echo "--- keycloak ---"
+docker logs --tail 80 han-chat-keycloak-1 2>&1 | grep -iE 'error|failed|exception|FATAL|OutOfMemory' || echo "(нет совпадений)"
+
+echo
+echo "=== 7. Память (Keycloak часто падает от OOM) ==="
+free -h
+docker stats --no-stream --format 'table {{.Name}}\t{{.MemUsage}}\t{{.CPUPerc}}' \
+  han-chat-keycloak-1 han-chat-api-backend-1 2>/dev/null || true
+
+echo
+echo "=== 8. Exited контейнеры (в метриках было exited=2) ==="
+docker ps -a --filter status=exited --format 'table {{.Names}}\t{{.Status}}\t{{.Image}}'

ops-monitoring/han-vm-metrics.sh

@@ -0,0 +1,345 @@
+#!/usr/bin/env bash
+# HAN Chat VM metrics snapshot
+# Host: Ubuntu 24.04 + Docker + UFW + fail2ban
+#
+# Usage:
+#   sudo ./han-vm-metrics.sh              # human report to stdout
+#   sudo ./han-vm-metrics.sh --json       # JSON to stdout
+#   sudo ./han-vm-metrics.sh --alert      # exit 1 if thresholds breached
+#   sudo ./han-vm-metrics.sh --log /var/log/han-vm-metrics.log
+#
+# Cron example (every 5 min):
+#   */5 * * * * root /opt/han-chat/ops/han-vm-metrics.sh --alert --log /var/log/han-vm-metrics.log
+#
+# Disk notes:
+#   Checks both / and /var/lib/docker. If Docker Root Dir is on a separate
+#   mount, its fill level is reported and alerted independently.
+
+set -euo pipefail
+
+MODE="text"
+LOG_FILE=""
+ALERT=0
+DISK_WARN_PCT="${DISK_WARN_PCT:-80}"
+DISK_CRIT_PCT="${DISK_CRIT_PCT:-90}"
+LOAD_WARN_MULT="${LOAD_WARN_MULT:-1.5}"   # vs CPU count
+MEM_WARN_PCT="${MEM_WARN_PCT:-90}"
+CERT_WARN_DAYS="${CERT_WARN_DAYS:-21}"
+SSH_FAIL_WARN="${SSH_FAIL_WARN:-50}"      # failed attempts last hour
+DOMAIN="${METRICS_DOMAIN:-chat.han0107.ru}"
+DOCKER_ROOT="${DOCKER_ROOT:-/var/lib/docker}"
+
+require_value() {
+  local opt="$1" val="${2:-}"
+  if [[ -z "$val" || "$val" == -* ]]; then
+    echo "Error: $opt requires a value" >&2
+    exit 2
+  fi
+}
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --json) MODE="json"; shift ;;
+    --alert) ALERT=1; shift ;;
+    --log)
+      require_value "$1" "${2:-}"
+      LOG_FILE="$2"
+      shift 2
+      ;;
+    --domain)
+      require_value "$1" "${2:-}"
+      DOMAIN="$2"
+      shift 2
+      ;;
+    -h|--help)
+      sed -n '2,18p' "$0"
+      exit 0
+      ;;
+    *) echo "Unknown arg: $1" >&2; exit 2 ;;
+  esac
+done
+
+ts_utc="$(date -u +%Y-%m-%dT%H:%M:%SZ)"
+hostname_f="$(hostname -f 2>/dev/null || hostname)"
+kernel="$(uname -r)"
+uptime_s="$(cut -d. -f1 /proc/uptime)"
+cpu_n="$(nproc)"
+loadavg="$(cut -d' ' -f1-3 /proc/loadavg)"
+load1="$(cut -d' ' -f1 /proc/loadavg)"
+
+# --- memory ---
+mem_total_kb="$(awk '/MemTotal:/ {print $2}' /proc/meminfo)"
+mem_avail_kb="$(awk '/MemAvailable:/ {print $2}' /proc/meminfo)"
+mem_used_kb=$((mem_total_kb - mem_avail_kb))
+mem_used_pct=$((mem_used_kb * 100 / mem_total_kb))
+swap_total_kb="$(awk '/SwapTotal:/ {print $2}' /proc/meminfo)"
+swap_free_kb="$(awk '/SwapFree:/ {print $2}' /proc/meminfo)"
+swap_used_kb=$((swap_total_kb - swap_free_kb))
+
+# --- disk helper: path -> total_k used_k avail_k used_pct ---
+read_df() {
+  local path="$1"
+  local line total used avail pct_raw
+  line="$(df -P "$path" 2>/dev/null | awk 'NR==2 {print $2,$3,$4,$5}')"
+  if [[ -z "$line" ]]; then
+    echo "0 0 0 0"
+    return
+  fi
+  read -r total used avail pct_raw <<<"$line"
+  echo "$total $used $avail ${pct_raw%%%}"
+}
+
+# --- disk root ---
+read -r disk_total_k disk_used_k disk_avail_k disk_used_pct <<<"$(read_df /)"
+
+# --- disk Docker root (may be same FS as / or a separate mount) ---
+docker_disk_same_as_root=1
+docker_disk_total_k=0
+docker_disk_used_k=0
+docker_disk_avail_k=0
+docker_disk_used_pct=0
+if [[ -d "$DOCKER_ROOT" ]]; then
+  read -r docker_disk_total_k docker_disk_used_k docker_disk_avail_k docker_disk_used_pct <<<"$(read_df "$DOCKER_ROOT")"
+  root_src="$(findmnt -n -o SOURCE / 2>/dev/null || true)"
+  docker_src="$(findmnt -n -o SOURCE --target "$DOCKER_ROOT" 2>/dev/null || true)"
+  if [[ -n "$root_src" && -n "$docker_src" && "$root_src" != "$docker_src" ]]; then
+    docker_disk_same_as_root=0
+  fi
+fi
+
+# --- reboot / updates (lightweight; avoid apt list --upgradable) ---
+reboot_required=0
+[[ -f /var/run/reboot-required ]] && reboot_required=1
+pending_pkgs=0
+pending_security=0
+if [[ -x /usr/lib/update-notifier/apt-check ]]; then
+  # stderr: "<total>;<security>"
+  apt_check_out="$(/usr/lib/update-notifier/apt-check 2>&1 || true)"
+  pending_pkgs="${apt_check_out%%;*}"
+  pending_security="${apt_check_out##*;}"
+elif [[ -f /var/lib/update-notifier/updates-available ]]; then
+  pending_pkgs="$(grep -oE '^[0-9]+' /var/lib/update-notifier/updates-available 2>/dev/null | head -1 || true)"
+fi
+[[ "$pending_pkgs" =~ ^[0-9]+$ ]] || pending_pkgs=0
+[[ "$pending_security" =~ ^[0-9]+$ ]] || pending_security=0
+
+# --- firewall ---
+ufw_status="unknown"
+if command -v ufw >/dev/null 2>&1; then
+  ufw_status="$(ufw status 2>/dev/null | awk 'NR==1 {print tolower($2)}')"
+fi
+
+# --- fail2ban ---
+f2b_active="inactive"
+f2b_jails=0
+f2b_banned_now=0
+f2b_banned_total=0
+f2b_failed_total=0
+if command -v fail2ban-client >/dev/null 2>&1; then
+  if systemctl is-active --quiet fail2ban 2>/dev/null; then
+    f2b_active="active"
+    f2b_status="$(fail2ban-client status 2>/dev/null || true)"
+    f2b_jails="$(echo "$f2b_status" | awk -F: '/Jail list/ {gsub(/ /,"",$2); n=split($2,a,","); print n; exit}')"
+    [[ -z "$f2b_jails" ]] && f2b_jails=0
+    if fail2ban-client status sshd >/dev/null 2>&1; then
+      sshd_st="$(fail2ban-client status sshd 2>/dev/null || true)"
+      f2b_banned_now="$(echo "$sshd_st" | awk -F: '/Currently banned/ {gsub(/ /,"",$2); print $2; exit}')"
+      f2b_banned_total="$(echo "$sshd_st" | awk -F: '/Total banned/ {gsub(/ /,"",$2); print $2; exit}')"
+      f2b_failed_total="$(echo "$sshd_st" | awk -F: '/Total failed/ {gsub(/ /,"",$2); print $2; exit}')"
+    fi
+  fi
+fi
+f2b_banned_now="${f2b_banned_now:-0}"
+f2b_banned_total="${f2b_banned_total:-0}"
+f2b_failed_total="${f2b_failed_total:-0}"
+
+# --- SSH auth last hour (single journalctl pass) ---
+ssh_failed_1h=0
+ssh_invalid_1h=0
+ssh_accepted_1h=0
+if command -v journalctl >/dev/null 2>&1; then
+  ssh_journal="$(journalctl -u ssh --since '1 hour ago' --no-pager 2>/dev/null || true)"
+  ssh_failed_1h="$(grep -cE 'Failed password|Failed publickey|Connection closed by authenticating' <<<"$ssh_journal" || true)"
+  ssh_invalid_1h="$(grep -c 'Invalid user' <<<"$ssh_journal" || true)"
+  ssh_accepted_1h="$(grep -c 'Accepted' <<<"$ssh_journal" || true)"
+fi
+
+# --- listening public ports ---
+listen_ports="$(ss -tuln 2>/dev/null | awk '/LISTEN/ && ($5 ~ /\*:|:0\.0\.0\.0:|\[::\]:/) {
+  split($5,a,":"); print a[length(a)]
+}' | sort -n | uniq | tr '\n' ',' | sed 's/,$//')"
+
+# --- docker ---
+docker_running=0
+docker_unhealthy=0
+docker_exited=0
+docker_names_unhealthy=""
+if command -v docker >/dev/null 2>&1 && docker info >/dev/null 2>&1; then
+  docker_running="$(docker ps -q 2>/dev/null | wc -l | tr -d ' ')"
+  docker_exited="$(docker ps -aq -f status=exited 2>/dev/null | wc -l | tr -d ' ')"
+  while IFS= read -r line; do
+    [[ -z "$line" ]] && continue
+    name="${line%% *}"
+    health="${line##* }"
+    if [[ "$health" == "unhealthy" ]]; then
+      docker_unhealthy=$((docker_unhealthy + 1))
+      docker_names_unhealthy+="${name},"
+    fi
+  done < <(docker ps --format '{{.Names}} {{.Status}}' 2>/dev/null | awk '{
+    h="unknown"
+    if ($0 ~ /unhealthy/) h="unhealthy"
+    else if ($0 ~ /\(healthy\)/) h="healthy"
+    else if ($0 ~ /Up/) h="up"
+    print $1, h
+  }')
+  docker_names_unhealthy="${docker_names_unhealthy%,}"
+fi
+
+# --- TLS cert days left (local nginx :443) ---
+cert_days=-1
+cert_cn=""
+if command -v openssl >/dev/null 2>&1; then
+  cert_pem="$(echo | openssl s_client -connect 127.0.0.1:443 -servername "$DOMAIN" 2>/dev/null | openssl x509 2>/dev/null || true)"
+  if [[ -n "$cert_pem" ]]; then
+    cert_cn="$(echo "$cert_pem" | openssl x509 -noout -subject 2>/dev/null | sed 's/.*CN *= *//')"
+    end_date="$(echo "$cert_pem" | openssl x509 -noout -enddate 2>/dev/null | cut -d= -f2)"
+    end_epoch="$(date -d "$end_date" +%s 2>/dev/null || true)"
+    now_epoch="$(date +%s)"
+    if [[ -n "${end_epoch:-}" ]]; then
+      cert_days=$(( (end_epoch - now_epoch) / 86400 ))
+    fi
+  fi
+fi
+
+# --- alerts ---
+alerts=()
+load_warn="$(awk -v l="$load1" -v n="$cpu_n" -v m="$LOAD_WARN_MULT" 'BEGIN { print (l > n*m) ? 1 : 0 }')"
+[[ "$load_warn" == "1" ]] && alerts+=("high_load:${load1}>${cpu_n}*${LOAD_WARN_MULT}")
+[[ "$mem_used_pct" -ge "$MEM_WARN_PCT" ]] && alerts+=("high_memory:${mem_used_pct}%")
+[[ "$disk_used_pct" -ge "$DISK_CRIT_PCT" ]] && alerts+=("disk_critical:${disk_used_pct}%")
+[[ "$disk_used_pct" -ge "$DISK_WARN_PCT" && "$disk_used_pct" -lt "$DISK_CRIT_PCT" ]] && alerts+=("disk_warn:${disk_used_pct}%")
+# Separate Docker mount: independent thresholds (avoid duplicate alerts when same FS)
+if [[ "$docker_disk_same_as_root" -eq 0 ]]; then
+  [[ "$docker_disk_used_pct" -ge "$DISK_CRIT_PCT" ]] && alerts+=("docker_disk_critical:${docker_disk_used_pct}%")
+  [[ "$docker_disk_used_pct" -ge "$DISK_WARN_PCT" && "$docker_disk_used_pct" -lt "$DISK_CRIT_PCT" ]] && alerts+=("docker_disk_warn:${docker_disk_used_pct}%")
+fi
+[[ "$reboot_required" -eq 1 ]] && alerts+=("reboot_required")
+[[ "$docker_unhealthy" -gt 0 ]] && alerts+=("docker_unhealthy:${docker_names_unhealthy}")
+[[ "$cert_days" -ge 0 && "$cert_days" -le "$CERT_WARN_DAYS" ]] && alerts+=("cert_expiring:${cert_days}d")
+[[ "$ssh_failed_1h" -ge "$SSH_FAIL_WARN" ]] && alerts+=("ssh_bruteforce_1h:${ssh_failed_1h}")
+[[ "$f2b_active" != "active" ]] && alerts+=("fail2ban_inactive")
+[[ "$ufw_status" != "active" ]] && alerts+=("ufw_inactive")
+
+alert_count=${#alerts[@]}
+alert_joined="$(IFS=','; echo "${alerts[*]:-}")"
+
+# KiB -> MiB (integer; avoids spawning awk)
+mb() { echo "$((${1:-0} / 1024))"; }
+
+render_text() {
+  local docker_disk_note="same_fs_as_root"
+  [[ "$docker_disk_same_as_root" -eq 0 ]] && docker_disk_note="separate_mount"
+  cat <<EOF
+=== HAN VM metrics @ ${ts_utc} ===
+host:           ${hostname_f}
+kernel:         ${kernel}
+uptime_sec:     ${uptime_s}
+cpu_cores:      ${cpu_n}
+loadavg:        ${loadavg}
+memory:         ${mem_used_pct}% used ($(mb "$mem_used_kb")/$(mb "$mem_total_kb") MiB), avail $(mb "$mem_avail_kb") MiB
+swap:           $(mb "$swap_used_kb")/$(mb "$swap_total_kb") MiB
+disk_root:      ${disk_used_pct}% used ($(mb "$disk_used_k")/$(mb "$disk_total_k") MiB), avail $(mb "$disk_avail_k") MiB
+disk_docker:    ${docker_disk_used_pct}% used ($(mb "$docker_disk_used_k")/$(mb "$docker_disk_total_k") MiB), avail $(mb "$docker_disk_avail_k") MiB [${docker_disk_note}: ${DOCKER_ROOT}]
+reboot_needed:  ${reboot_required}
+apt_upgradable: ${pending_pkgs} (security=${pending_security})
+ufw:            ${ufw_status}
+listen_ports:   ${listen_ports:-none}
+fail2ban:       ${f2b_active}, jails=${f2b_jails}, banned_now=${f2b_banned_now}, banned_total=${f2b_banned_total}, failed_total=${f2b_failed_total}
+ssh_1h:         failed=${ssh_failed_1h}, invalid_user=${ssh_invalid_1h}, accepted=${ssh_accepted_1h}
+docker:         running=${docker_running}, exited=${docker_exited}, unhealthy=${docker_unhealthy}${docker_names_unhealthy:+ ($docker_names_unhealthy)}
+tls:            domain=${DOMAIN}, cn=${cert_cn:-n/a}, days_left=${cert_days}
+alerts(${alert_count}): ${alert_joined:-none}
+EOF
+}
+
+render_json() {
+  python3 - "$ts_utc" <<'PY' || true
+import json, os, sys
+ts = sys.argv[1]
+data = {
+  "ts": ts,
+  "host": os.environ["M_HOST"],
+  "kernel": os.environ["M_KERNEL"],
+  "uptime_sec": int(os.environ["M_UPTIME"]),
+  "cpu_cores": int(os.environ["M_CPU"]),
+  "loadavg": os.environ["M_LOADAVG"],
+  "load1": float(os.environ["M_LOAD1"]),
+  "memory_used_pct": int(os.environ["M_MEM_PCT"]),
+  "disk_root_used_pct": int(os.environ["M_DISK_PCT"]),
+  "disk_docker": {
+    "path": os.environ["M_D_DISK_PATH"],
+    "used_pct": int(os.environ["M_D_DISK_PCT"]),
+    "same_fs_as_root": os.environ["M_D_DISK_SAME"] == "1",
+  },
+  "reboot_required": os.environ["M_REBOOT"] == "1",
+  "apt_upgradable": int(os.environ["M_APT"]),
+  "apt_security": int(os.environ["M_APT_SEC"]),
+  "ufw": os.environ["M_UFW"],
+  "listen_ports": [p for p in os.environ.get("M_PORTS","").split(",") if p],
+  "fail2ban": {
+    "active": os.environ["M_F2B"] == "active",
+    "jails": int(os.environ["M_F2B_JAILS"]),
+    "banned_now": int(os.environ["M_F2B_NOW"]),
+    "banned_total": int(os.environ["M_F2B_TOT"]),
+    "failed_total": int(os.environ["M_F2B_FAIL"]),
+  },
+  "ssh_1h": {
+    "failed": int(os.environ["M_SSH_FAIL"]),
+    "invalid_user": int(os.environ["M_SSH_INV"]),
+    "accepted": int(os.environ["M_SSH_OK"]),
+  },
+  "docker": {
+    "running": int(os.environ["M_D_RUN"]),
+    "exited": int(os.environ["M_D_EXIT"]),
+    "unhealthy": int(os.environ["M_D_BAD"]),
+    "unhealthy_names": [n for n in os.environ.get("M_D_BAD_N","").split(",") if n],
+  },
+  "tls": {
+    "domain": os.environ["M_DOMAIN"],
+    "cn": os.environ.get("M_CERT_CN") or None,
+    "days_left": int(os.environ["M_CERT_DAYS"]),
+  },
+  "alerts": [a for a in os.environ.get("M_ALERTS","").split(",") if a],
+}
+print(json.dumps(data, ensure_ascii=False))
+PY
+}
+
+export M_HOST="$hostname_f" M_KERNEL="$kernel" M_UPTIME="$uptime_s" M_CPU="$cpu_n"
+export M_LOADAVG="$loadavg" M_LOAD1="$load1" M_MEM_PCT="$mem_used_pct" M_DISK_PCT="$disk_used_pct"
+export M_D_DISK_PATH="$DOCKER_ROOT" M_D_DISK_PCT="$docker_disk_used_pct" M_D_DISK_SAME="$docker_disk_same_as_root"
+export M_REBOOT="$reboot_required" M_APT="$pending_pkgs" M_APT_SEC="$pending_security"
+export M_UFW="$ufw_status" M_PORTS="$listen_ports"
+export M_F2B="$f2b_active" M_F2B_JAILS="$f2b_jails" M_F2B_NOW="$f2b_banned_now"
+export M_F2B_TOT="$f2b_banned_total" M_F2B_FAIL="$f2b_failed_total"
+export M_SSH_FAIL="$ssh_failed_1h" M_SSH_INV="$ssh_invalid_1h" M_SSH_OK="$ssh_accepted_1h"
+export M_D_RUN="$docker_running" M_D_EXIT="$docker_exited" M_D_BAD="$docker_unhealthy" M_D_BAD_N="$docker_names_unhealthy"
+export M_DOMAIN="$DOMAIN" M_CERT_CN="$cert_cn" M_CERT_DAYS="$cert_days" M_ALERTS="$alert_joined"
+
+if [[ "$MODE" == "json" ]]; then
+  out="$(render_json)"
+else
+  out="$(render_text)"
+fi
+
+printf '%s\n' "$out"
+
+if [[ -n "$LOG_FILE" ]]; then
+  mkdir -p "$(dirname "$LOG_FILE")"
+  printf '%s\n' "$out" >>"$LOG_FILE"
+fi
+
+if [[ "$ALERT" -eq 1 && "$alert_count" -gt 0 ]]; then
+  exit 1
+fi
+exit 0

ops-monitoring/pg_connection_test.sh

@@ -0,0 +1,36 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# 1. Безопасная загрузка .env: убираем комментарии, удаляем \r (Windows-переносы) и делаем export
+export $(grep -v '^#' .env | sed 's/\r$//' | xargs)
+
+echo "=== 1. Параметры из .env ==="
+grep -E '^(HAN_PG_HOST|HAN_PG_PORT|KEYCLOAK_DB_)' .env | grep -v PASSWORD || echo "Переменные не найдены!"
+
+# 2. Проверка TCP (используем переменные, а не хардкод)
+echo -e "\n=== 2. Проверка TCP (${HAN_PG_HOST}:${HAN_PG_PORT}) ==="
+if command -v nc &> /dev/null; then
+    nc -zv "${HAN_PG_HOST}" "${HAN_PG_PORT}" || echo "❌ Не удалось подключиться по TCP"
+else
+    echo "⚠️ Утилита nc не найдена, пропускаем проверку TCP"
+fi
+
+# 3. Проверка подключения psql
+echo -e "\n=== 3. Проверка подключения к БД ==="
+
+# ВАЖНО: Передаем пароль через переменную окружения, чтобы скрипт не зависал!
+# Если KEYCLOAK_DB_PASSWORD пуст, psql все равно спросит его вручную.
+export PGPASSWORD="${KEYCLOAK_DB_PASSWORD:-}"
+
+# Проверяем, установлен ли psql
+if ! command -v psql &> /dev/null; then
+    echo "❌ Утилита psql не найдена на хост-машине. Установите postgresql-client."
+    exit 1
+fi
+
+# Выполняем запрос. Используем переменную порта и добавляем fallback для sslrootcert, если он не задан
+psql "host=${HAN_PG_HOST} port=${HAN_PG_PORT} dbname=han_chat user=${KEYCLOAK_DB_USERNAME} sslmode=verify-full sslrootcert=${PG_CA_HOST_PATH:-}" \
+  -c "SELECT current_user, current_schema(); SHOW search_path;"
+
+# Очищаем пароль из окружения после выполнения (хорошая практика)
+unset PGPASSWORD