Files
portal/RUNBOOK.md
T
Дмитрий b241c79773 docs: add RUNBOOK.md — production deployment runbook
Audit #2 Phase 14 P2 fix. Covers: system requirements, DB setup
(ICU collation + roles + migrations + grants), partition bootstrap,
frontend build, Supervisor queue config, cron scheduler, Nginx,
health checks, rolling update sequence, rollback, dev seed,
common issues. cspell-words.txt +mbstring +pcntl (PHP ext names).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-13 13:35:19 +03:00

7.6 KiB

Лидерра CRM — Production Deployment Runbook

Version: 1.0 от 2026-05-13
Stack: Laravel 13 · Vue 3 + Vuetify 3 · PostgreSQL 16 · Redis 7 · PHP 8.3
Cloud: Yandex Cloud, region ru-central1


1. System Requirements

Component Version Notes
PHP 8.3+ Extensions: pdo_pgsql, pgsql, redis, bcmath, mbstring, openssl, tokenizer, xml, ctype, fileinfo, pcntl
PostgreSQL 16 ICU collation support required (--with-icu compile flag)
Redis 7.x Sessions, queues, cache
Node.js 20+ Frontend build only
Composer 2.x
Supervisor 4.x Queue worker process management

2. Environment Configuration

Copy .env.example to .env and set all required values:

cp app/.env.example app/.env

Critical variables:

APP_ENV=production
APP_KEY=                        # php artisan key:generate
APP_URL=https://crm.example.com

DB_CONNECTION=pgsql
DB_HOST=127.0.0.1
DB_PORT=5432
DB_DATABASE=liderra
DB_USERNAME=crm_migrator        # migration role — full DDL rights
DB_PASSWORD=<secret>

REDIS_HOST=127.0.0.1
REDIS_PORT=6379
REDIS_PASSWORD=<secret>

QUEUE_CONNECTION=redis
SESSION_DRIVER=redis
CACHE_STORE=redis

MAIL_MAILER=smtp
MAIL_HOST=smtp.unisender.com
MAIL_PORT=587
MAIL_USERNAME=<unisender-go-api-key>
MAIL_ENCRYPTION=tls
MAIL_FROM_ADDRESS=noreply@liderra.ru
MAIL_FROM_NAME="Лидерра"

3. Database Setup

3.1 Create database with ICU collation

-- Run as PostgreSQL superuser
CREATE DATABASE liderra
    ENCODING 'UTF8'
    LOCALE_PROVIDER 'icu'
    ICU_LOCALE 'ru-RU'
    TEMPLATE template0;

3.2 Create application roles

# Run as PostgreSQL superuser
psql -U postgres liderra < db/00_create_roles.sql

The script creates 5 roles: crm_app_user, crm_app_admin, crm_readonly, crm_migrator, crm_supplier_worker (BYPASSRLS).

3.3 Run migrations

cd app
php artisan migrate --force

This loads db/schema.sql (v8.19+) via the single bootstrap migration load_initial_schema.php.

3.4 Apply grants

psql -U postgres liderra < db/02_grants.sql

3.5 Create initial partition tables

Partitioned tables (lead_audit_log, supplier_session_log, etc.) require month-based child partitions to exist before any inserts:

cd app
php artisan partitions:create-months

Run this once after migration. The scheduler maintains partitions automatically thereafter (see §7).

3.6 Apply pg_audit extension (pre-prod)

-- Run as PostgreSQL superuser
CREATE EXTENSION IF NOT EXISTS pgaudit;

4. Application Bootstrap

cd app

# Install PHP dependencies
composer install --no-dev --optimize-autoloader

# Generate app key (first deploy only)
php artisan key:generate --force

# Clear and cache config/routes/views
php artisan config:cache
php artisan route:cache
php artisan view:cache

# Run storage symlink
php artisan storage:link

5. Frontend Build

cd app
npm ci
npm run build

Output goes to app/public/build/. Confirm app/public/build/manifest.json exists.


6. Queue Worker (Supervisor)

Create /etc/supervisor/conf.d/liderra-worker.conf:

[program:liderra-worker]
process_name=%(program_name)s_%(process_num)02d
command=php /path/to/app/artisan queue:work redis --sleep=3 --tries=3 --max-time=3600
autostart=true
autorestart=true
stopasgroup=true
killasgroup=true
user=www-data
numprocs=2
redirect_stderr=true
stdout_logfile=/var/log/liderra-worker.log
stopwaitsecs=3600
supervisorctl reread
supervisorctl update
supervisorctl start liderra-worker:*

7. Scheduler (Cron)

Add to the deployment user's crontab (crontab -e):

* * * * * cd /path/to/app && php artisan schedule:run >> /dev/null 2>&1

The scheduler runs these jobs automatically:

Command/Job Schedule Purpose
projects:reset-delivered-today daily 00:00 MSK Reset daily lead counter
projects:reset-monthly 1st of month 00:00 MSK Reset monthly counter for tier lookup
partitions:create-months daily Create PostgreSQL partition tables for upcoming months
RefreshSupplierSessionJob hourly + 20:15 MSK Supplier API session tokens
SyncSupplierProjectsJob 20:30 MSK daily Sync supplier project list
CleanupInactiveSupplierProjectsJob 02:00 MSK daily Archive stale supplier projects
supplier:retry-failed hourly Retry failed supplier lead deliveries
CsvReconcileJob hourly CSV reconciliation (reserve lead intake channel)

8. Web Server (Nginx)

server {
    listen 443 ssl;
    server_name crm.example.com;

    root /path/to/app/public;
    index index.php;

    location / {
        try_files $uri $uri/ /index.php?$query_string;
    }

    location ~ \.php$ {
        fastcgi_pass unix:/var/run/php/php8.3-fpm.sock;
        fastcgi_index index.php;
        fastcgi_param SCRIPT_FILENAME $realpath_root$fastcgi_script_name;
        include fastcgi_params;
    }

    location ~* \.(js|css|png|jpg|svg|woff2)$ {
        expires 1y;
        add_header Cache-Control "public, immutable";
    }
}

9. Health Checks

# PHP and Laravel bootstrap
cd app && php artisan about

# Database connection
php artisan db:show

# Scheduler registered entries
php artisan schedule:list

# Queue worker status
supervisorctl status liderra-worker:*

# Redis connection
redis-cli -h 127.0.0.1 ping

# Application HTTP
curl -I https://crm.example.com/login

Expected responses:

  • php artisan about — no errors, APP_ENV=production
  • php artisan schedule:list — 9 entries including partitions:create-months
  • Redis: PONG
  • HTTP /login: 200

10. Deployment Sequence (Rolling Update)

# 1. Pull latest code
git pull origin main

# 2. Install/update dependencies
cd app && composer install --no-dev --optimize-autoloader
npm ci && npm run build

# 3. Run new migrations (if any)
php artisan migrate --force

# 4. Recache configuration
php artisan config:cache
php artisan route:cache
php artisan view:cache

# 5. Restart queue workers (pick up new code)
supervisorctl restart liderra-worker:*

11. Rollback

# Revert to previous release tag
git checkout <previous-tag>
cd app
composer install --no-dev --optimize-autoloader
npm ci && npm run build
php artisan migrate:rollback   # only if the migration is reversible
php artisan config:cache
supervisorctl restart liderra-worker:*

Warning: Schema migrations are not always reversible. Always take a PostgreSQL dump before deploying schema changes.

pg_dump -U crm_migrator liderra > backup_$(date +%Y%m%d_%H%M%S).sql

12. Development Seed

For staging/dev environments only:

cd app
php artisan db:seed --class=DemoSeeder --force

Creates: admin@demo.local / password, 3 projects, 14 demo deals.

Never run DemoSeeder on production.


13. Common Issues

Symptom Likely Cause Fix
SQLSTATE[08006] on boot Wrong DB_HOST (use 127.0.0.1, not localhost on Windows) Set DB_HOST=127.0.0.1
Partition insert error on new month partitions:create-months not run php artisan partitions:create-months
Queue jobs not processing Supervisor not running or wrong user supervisorctl status; check stdout_logfile
CSS/JS 404 after deploy Frontend not rebuilt or storage:link missing npm run build + php artisan storage:link
jwt expired from supplier API Supplier session not refreshed php artisan tinkerdispatch(new RefreshSupplierSessionJob)
Scheduler not running Cron not set up Verify crontab entry; php artisan schedule:run --verbose