feat: Version 3.5.2 - Configuration Stripe et gestion des immeubles
- Configuration complète Stripe pour les 3 environnements (DEV/REC/PROD) * DEV: Clés TEST Pierre (mode test) * REC: Clés TEST Client (mode test) * PROD: Clés LIVE Client (mode live) - Ajout de la gestion des bases de données immeubles/bâtiments * Configuration buildings_database pour DEV/REC/PROD * Service BuildingService pour enrichissement des adresses - Optimisations pages et améliorations ergonomie - Mises à jour des dépendances Composer - Nettoyage des fichiers obsolètes 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
pierre
608
MONITORING.md
Normal file
608
MONITORING.md
Normal file
@@ -0,0 +1,608 @@
|
||||
# Guide de Monitoring - Container Incus (Application Web PHP + MariaDB)
|
||||
|
||||
## Vue d'ensemble
|
||||
|
||||
Ce guide décrit les métriques essentielles à surveiller pour un container Incus hébergeant une application web PHP avec API mobile et base de données MariaDB.
|
||||
|
||||
---
|
||||
|
||||
## 1. Ressources Système
|
||||
|
||||
### CPU
|
||||
**Pourquoi ?** Identifier les pics de charge et les processus gourmands
|
||||
|
||||
```bash
|
||||
# Utilisation CPU globale du container
|
||||
incus info nx4 | grep "CPU utilisé"
|
||||
|
||||
# Détail par processus (dans le container)
|
||||
top -bn1 | head -20
|
||||
htop
|
||||
```
|
||||
|
||||
**Métriques à surveiller :**
|
||||
- Load average (devrait être < nombre de CPU)
|
||||
- % CPU par processus (MariaDB, PHP-FPM, nginx)
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : Load average > 70% des CPU
|
||||
- 🚨 Critical : Load average > 150% des CPU pendant >5min
|
||||
|
||||
---
|
||||
|
||||
### Mémoire RAM
|
||||
**Pourquoi ?** Éviter les OOM (Out of Memory) qui tuent les processus
|
||||
|
||||
```bash
|
||||
# Vue globale depuis le host
|
||||
incus info nx4 | grep "Mémoire"
|
||||
|
||||
# Détail dans le container
|
||||
free -h
|
||||
ps aux --sort=-%mem | head -10
|
||||
```
|
||||
|
||||
**Métriques à surveiller :**
|
||||
- RAM utilisée / totale
|
||||
- Swap utilisé (devrait rester minimal)
|
||||
- Top processus consommateurs
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : RAM > 85%
|
||||
- 🚨 Critical : RAM > 95% ou Swap > 1GB
|
||||
|
||||
---
|
||||
|
||||
### Disque I/O
|
||||
**Pourquoi ?** MariaDB est très sensible aux lenteurs disque
|
||||
|
||||
```bash
|
||||
# Sur le host
|
||||
iostat -x 2 5
|
||||
|
||||
# Dans le container
|
||||
iostat -x 1 3
|
||||
iotop -oa
|
||||
```
|
||||
|
||||
**Métriques à surveiller :**
|
||||
- Latence disque (await)
|
||||
- IOPS (r/s, w/s)
|
||||
- % utilisation disque
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : await > 50ms
|
||||
- 🚨 Critical : await > 200ms ou %util > 90%
|
||||
|
||||
---
|
||||
|
||||
### Espace Disque
|
||||
**Pourquoi ?** MariaDB ne peut plus écrire si disque plein
|
||||
|
||||
```bash
|
||||
df -h
|
||||
du -sh /var/lib/mysql
|
||||
```
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : > 80% utilisé
|
||||
- 🚨 Critical : > 90% utilisé
|
||||
|
||||
---
|
||||
|
||||
## 2. PHP-FPM (Application Web)
|
||||
|
||||
### Pool de Workers
|
||||
**Pourquoi ?** Cause #1 des timeouts et coupures de service (votre cas !)
|
||||
|
||||
```bash
|
||||
# Nombre de workers actifs
|
||||
ps aux | grep "php-fpm: pool www" | wc -l
|
||||
|
||||
# Config du pool
|
||||
grep "^pm" /etc/php/8.3/fpm/pool.d/www.conf
|
||||
|
||||
# Logs d'alertes
|
||||
tail -f /var/log/php8.3-fpm.log | grep "max_children"
|
||||
```
|
||||
|
||||
**Métriques critiques :**
|
||||
- Nombre de workers actifs vs `pm.max_children`
|
||||
- Warnings "max_children reached"
|
||||
- Slow requests (>2s)
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : Workers actifs > 80% de max_children
|
||||
- 🚨 Critical : Warning "max_children reached" apparaît
|
||||
|
||||
**Configuration recommandée pour votre cas :**
|
||||
```ini
|
||||
pm = dynamic
|
||||
pm.max_children = 50-100 (selon RAM disponible)
|
||||
pm.start_servers = 10-20
|
||||
pm.min_spare_servers = 5-10
|
||||
pm.max_spare_servers = 20-35
|
||||
pm.max_requests = 500
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### Temps de Réponse PHP
|
||||
**Pourquoi ?** Scripts lents = workers bloqués
|
||||
|
||||
```bash
|
||||
# Activer le slow log PHP-FPM
|
||||
# Dans /etc/php/8.3/fpm/pool.d/www.conf
|
||||
slowlog = /var/log/php8.3-fpm-slow.log
|
||||
request_slowlog_timeout = 3s
|
||||
|
||||
# Voir les requêtes lentes
|
||||
tail -f /var/log/php8.3-fpm-slow.log
|
||||
```
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : Requêtes > 3s
|
||||
- 🚨 Critical : Requêtes > 10s ou timeouts fréquents
|
||||
|
||||
---
|
||||
|
||||
## 3. MariaDB / MySQL
|
||||
|
||||
### Connexions
|
||||
**Pourquoi ?** Trop de connexions = refus de nouvelles connexions
|
||||
|
||||
```bash
|
||||
mysql -e "SHOW STATUS LIKE 'Threads_connected';"
|
||||
mysql -e "SHOW STATUS LIKE 'Max_used_connections';"
|
||||
mysql -e "SHOW VARIABLES LIKE 'max_connections';"
|
||||
mysql -e "SHOW FULL PROCESSLIST;"
|
||||
```
|
||||
|
||||
**Métriques critiques :**
|
||||
- Connexions actives vs max_connections
|
||||
- Connexions en attente / bloquées
|
||||
- Requêtes longues (>5s)
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : Connexions > 80% de max_connections
|
||||
- 🚨 Critical : Connexions = max_connections
|
||||
|
||||
**Config recommandée :**
|
||||
```ini
|
||||
max_connections = 200-500 (selon votre trafic)
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### Slow Queries
|
||||
**Pourquoi ?** Requêtes lentes = workers PHP bloqués
|
||||
|
||||
```bash
|
||||
# Activer le slow query log
|
||||
mysql -e "SET GLOBAL slow_query_log = 'ON';"
|
||||
mysql -e "SET GLOBAL long_query_time = 2;"
|
||||
mysql -e "SET GLOBAL log_queries_not_using_indexes = 'ON';"
|
||||
|
||||
# Voir les slow queries
|
||||
tail -f /var/lib/mysql/slow-query.log
|
||||
# ou
|
||||
mysqldumpslow /var/lib/mysql/slow-query.log
|
||||
```
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : Requêtes > 2s
|
||||
- 🚨 Critical : Requêtes > 10s ou >50 slow queries/min
|
||||
|
||||
---
|
||||
|
||||
### InnoDB Buffer Pool
|
||||
**Pourquoi ?** Si trop petit, beaucoup de lectures disque (lent)
|
||||
|
||||
```bash
|
||||
mysql -e "SHOW VARIABLES LIKE 'innodb_buffer_pool_size';"
|
||||
mysql -e "SHOW STATUS LIKE 'Innodb_buffer_pool_read%';"
|
||||
```
|
||||
|
||||
**Métriques critiques :**
|
||||
- Buffer pool hit ratio (devrait être >99%)
|
||||
- Read requests vs reads from disk
|
||||
|
||||
**Config recommandée :**
|
||||
- `innodb_buffer_pool_size` = 70-80% de la RAM dédiée à MySQL
|
||||
- Pour 20GB de données : `innodb_buffer_pool_size = 16G`
|
||||
|
||||
---
|
||||
|
||||
### Deadlocks et Locks
|
||||
**Pourquoi ?** Peuvent causer des timeouts
|
||||
|
||||
```bash
|
||||
mysql -e "SHOW ENGINE INNODB STATUS\G" | grep -A 50 "LATEST DETECTED DEADLOCK"
|
||||
mysql -e "SHOW OPEN TABLES WHERE In_use > 0;"
|
||||
```
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : >1 deadlock/heure
|
||||
- 🚨 Critical : Tables lockées >30s
|
||||
|
||||
---
|
||||
|
||||
## 4. Nginx / Serveur Web
|
||||
|
||||
### Connexions et Erreurs
|
||||
**Pourquoi ?** Identifier les 502/504 (backend timeout)
|
||||
|
||||
```bash
|
||||
# Connexions actives
|
||||
netstat -an | grep :80 | wc -l
|
||||
netstat -an | grep :443 | wc -l
|
||||
|
||||
# Erreurs récentes
|
||||
tail -100 /var/log/nginx/error.log | grep -E "502|504|timeout"
|
||||
tail -100 /var/log/nginx/access.log | grep -E " 502 | 504 "
|
||||
```
|
||||
|
||||
**Métriques critiques :**
|
||||
- Erreurs 502 Bad Gateway (PHP-FPM down)
|
||||
- Erreurs 504 Gateway Timeout (PHP-FPM trop lent)
|
||||
- Connexions actives
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : >5 erreurs 502/504 en 5min
|
||||
- 🚨 Critical : >20 erreurs 502/504 en 5min
|
||||
|
||||
---
|
||||
|
||||
### Worker Connections
|
||||
**Pourquoi ?** Limite le nombre de connexions simultanées
|
||||
|
||||
```bash
|
||||
# Config nginx
|
||||
grep worker_connections /etc/nginx/nginx.conf
|
||||
ps aux | grep nginx | wc -l
|
||||
```
|
||||
|
||||
**Config recommandée :**
|
||||
```nginx
|
||||
worker_processes auto;
|
||||
worker_connections 2048;
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Réseau
|
||||
|
||||
### Bande Passante
|
||||
**Pourquoi ?** Identifier les pics de trafic
|
||||
|
||||
```bash
|
||||
# Depuis le host
|
||||
incus info nx4 | grep -A 10 "eth0:"
|
||||
|
||||
# Dans le container
|
||||
iftop -i eth0
|
||||
vnstat -i eth0
|
||||
```
|
||||
|
||||
**Métriques :**
|
||||
- Octets reçus/émis
|
||||
- Paquets reçus/émis
|
||||
- Erreurs réseau
|
||||
|
||||
---
|
||||
|
||||
### Connexions TCP
|
||||
**Pourquoi ?** Trop de connexions TIME_WAIT = problème
|
||||
|
||||
```bash
|
||||
netstat -an | grep -E "ESTABLISHED|TIME_WAIT|CLOSE_WAIT" | wc -l
|
||||
ss -s
|
||||
```
|
||||
|
||||
**Seuils d'alerte :**
|
||||
- ⚠️ Warning : >1000 connexions TIME_WAIT
|
||||
- 🚨 Critical : >5000 connexions TIME_WAIT
|
||||
|
||||
---
|
||||
|
||||
## 6. Logs et Événements
|
||||
|
||||
### Logs Système
|
||||
```bash
|
||||
# Container Debian
|
||||
journalctl -u nginx -n 100
|
||||
journalctl -u php8.3-fpm -n 100
|
||||
journalctl -u mariadb -n 100
|
||||
|
||||
# Container Alpine (si pas de systemd)
|
||||
tail -100 /var/log/messages
|
||||
```
|
||||
|
||||
### Logs Applicatifs
|
||||
```bash
|
||||
# PHP errors
|
||||
tail -100 /var/log/php8.3-fpm.log
|
||||
tail -100 /var/www/html/logs/*.log
|
||||
|
||||
# Nginx
|
||||
tail -100 /var/log/nginx/error.log
|
||||
tail -100 /var/log/nginx/access.log
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 7. Scripts de Monitoring Automatisés
|
||||
|
||||
### Script de Monitoring Global
|
||||
|
||||
Créer `/root/monitor.sh` :
|
||||
|
||||
```bash
|
||||
#!/bin/bash
|
||||
|
||||
LOG_FILE="/var/log/system-monitor.log"
|
||||
ALERT_FILE="/var/log/alerts.log"
|
||||
|
||||
echo "=== Monitoring $(date) ===" >> $LOG_FILE
|
||||
|
||||
# 1. CPU
|
||||
LOAD=$(uptime | awk -F'load average:' '{print $2}' | awk '{print $1}')
|
||||
echo "Load: $LOAD" >> $LOG_FILE
|
||||
|
||||
# 2. RAM
|
||||
RAM_PERCENT=$(free | awk '/Mem:/ {printf("%.1f"), ($3/$2)*100}')
|
||||
echo "RAM: ${RAM_PERCENT}%" >> $LOG_FILE
|
||||
if (( $(echo "$RAM_PERCENT > 85" | bc -l) )); then
|
||||
echo "[WARNING] RAM > 85%: ${RAM_PERCENT}%" >> $ALERT_FILE
|
||||
fi
|
||||
|
||||
# 3. PHP-FPM Workers
|
||||
PHP_WORKERS=$(ps aux | grep "php-fpm: pool www" | wc -l)
|
||||
PHP_MAX=$(grep "^pm.max_children" /etc/php/8.3/fpm/pool.d/www.conf | awk '{print $3}')
|
||||
echo "PHP Workers: $PHP_WORKERS / $PHP_MAX" >> $LOG_FILE
|
||||
if [ $PHP_WORKERS -gt $(echo "$PHP_MAX * 0.8" | bc) ]; then
|
||||
echo "[WARNING] PHP Workers > 80%: $PHP_WORKERS / $PHP_MAX" >> $ALERT_FILE
|
||||
fi
|
||||
|
||||
# 4. MySQL Connexions
|
||||
MYSQL_CONN=$(mysql -e "SHOW STATUS LIKE 'Threads_connected';" | awk 'NR==2 {print $2}')
|
||||
MYSQL_MAX=$(mysql -e "SHOW VARIABLES LIKE 'max_connections';" | awk 'NR==2 {print $2}')
|
||||
echo "MySQL Connections: $MYSQL_CONN / $MYSQL_MAX" >> $LOG_FILE
|
||||
if [ $MYSQL_CONN -gt $(echo "$MYSQL_MAX * 0.8" | bc) ]; then
|
||||
echo "[WARNING] MySQL Connections > 80%: $MYSQL_CONN / $MYSQL_MAX" >> $ALERT_FILE
|
||||
fi
|
||||
|
||||
# 5. Disque
|
||||
DISK_PERCENT=$(df -h / | awk 'NR==2 {print $5}' | sed 's/%//')
|
||||
echo "Disk: ${DISK_PERCENT}%" >> $LOG_FILE
|
||||
if [ $DISK_PERCENT -gt 80 ]; then
|
||||
echo "[WARNING] Disk > 80%: ${DISK_PERCENT}%" >> $ALERT_FILE
|
||||
fi
|
||||
|
||||
# 6. Erreurs nginx
|
||||
NGINX_ERRORS=$(grep -c "error" /var/log/nginx/error.log | tail -100)
|
||||
if [ $NGINX_ERRORS -gt 10 ]; then
|
||||
echo "[WARNING] Nginx errors: $NGINX_ERRORS in last 100 lines" >> $ALERT_FILE
|
||||
fi
|
||||
|
||||
echo "" >> $LOG_FILE
|
||||
```
|
||||
|
||||
**Installation :**
|
||||
```bash
|
||||
chmod +x /root/monitor.sh
|
||||
|
||||
# Exécuter toutes les 5 minutes
|
||||
(crontab -l 2>/dev/null; echo "*/5 * * * * /root/monitor.sh") | crontab -
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### Script de Monitoring PHP-FPM Spécifique
|
||||
|
||||
Créer `/root/monitor-php-fpm.sh` :
|
||||
|
||||
```bash
|
||||
#!/bin/bash
|
||||
|
||||
LOG="/var/log/php-fpm-monitor.log"
|
||||
TIMESTAMP=$(date '+%Y-%m-%d %H:%M:%S')
|
||||
|
||||
# Compter les workers (sans grep lui-même)
|
||||
WORKERS=$(ps aux | grep 'php-fpm: pool www' | grep -v grep | wc -l)
|
||||
MAX_CHILDREN=$(grep '^pm.max_children' /etc/php/8.3/fpm/pool.d/www.conf | awk '{print $3}')
|
||||
PERCENT=$(echo "scale=1; ($WORKERS / $MAX_CHILDREN) * 100" | bc)
|
||||
|
||||
# Log au format: timestamp,workers,max,percentage
|
||||
echo "$TIMESTAMP,$WORKERS,$MAX_CHILDREN,$PERCENT%" >> $LOG
|
||||
|
||||
# Alerte si >80%
|
||||
if (( $(echo "$PERCENT > 80" | bc -l) )); then
|
||||
echo "[$TIMESTAMP] [ALERT] PHP Workers > 80%: $WORKERS / $MAX_CHILDREN ($PERCENT%)" >> /var/log/alerts.log
|
||||
fi
|
||||
|
||||
# Vérifier si max_children atteint dans les logs récents
|
||||
if tail -10 /var/log/php8.3-fpm.log | grep -q "max_children"; then
|
||||
echo "[$TIMESTAMP] [CRITICAL] MAX_CHILDREN REACHED!" >> /var/log/alerts.log
|
||||
tail -3 /var/log/php8.3-fpm.log >> /var/log/alerts.log
|
||||
fi
|
||||
```
|
||||
|
||||
**Installation :**
|
||||
```bash
|
||||
chmod +x /root/monitor-php-fpm.sh
|
||||
|
||||
# Exécuter toutes les minutes
|
||||
(crontab -l 2>/dev/null; echo "* * * * * /root/monitor-php-fpm.sh") | crontab -
|
||||
```
|
||||
|
||||
**Visualiser les données :**
|
||||
```bash
|
||||
# Afficher les dernières 60 minutes
|
||||
tail -60 /var/log/php-fpm-monitor.log
|
||||
|
||||
# Voir l'évolution graphique (nécessite gnuplot)
|
||||
echo 'set datafile separator ","; plot "/var/log/php-fpm-monitor.log" using 2 with lines title "Workers"' | gnuplot -p
|
||||
|
||||
# Statistiques rapides
|
||||
echo "Max workers last hour: $(tail -60 /var/log/php-fpm-monitor.log | cut -d',' -f2 | sort -n | tail -1)"
|
||||
echo "Min workers last hour: $(tail -60 /var/log/php-fpm-monitor.log | cut -d',' -f2 | sort -n | head -1)"
|
||||
echo "Avg workers last hour: $(tail -60 /var/log/php-fpm-monitor.log | cut -d',' -f2 | awk '{sum+=$1} END {print sum/NR}')"
|
||||
|
||||
# Alertes récentes
|
||||
tail -20 /var/log/alerts.log | grep "PHP Workers"
|
||||
```
|
||||
|
||||
**Rotation automatique des logs :**
|
||||
Créer `/etc/logrotate.d/php-fpm-monitor` :
|
||||
```
|
||||
/var/log/php-fpm-monitor.log {
|
||||
daily
|
||||
rotate 30
|
||||
compress
|
||||
missingok
|
||||
notifempty
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 8. Solutions de Monitoring Automatisées
|
||||
|
||||
### Option 1 : Netdata (Recommandé)
|
||||
**Avantages :** Installation simple, interface web, détection auto des services
|
||||
|
||||
```bash
|
||||
# Installation
|
||||
bash <(curl -Ss https://get.netdata.cloud/kickstart.sh) --dont-wait
|
||||
|
||||
# Accessible sur http://IP:19999
|
||||
# Détecte automatiquement : PHP-FPM, MariaDB, Nginx, ressources système
|
||||
```
|
||||
|
||||
**Métriques auto-détectées :**
|
||||
- ✅ CPU, RAM, I/O, réseau
|
||||
- ✅ PHP-FPM (workers, slow requests)
|
||||
- ✅ MariaDB (connexions, queries, locks)
|
||||
- ✅ Nginx (connexions, erreurs)
|
||||
|
||||
---
|
||||
|
||||
### Option 2 : Prometheus + Grafana
|
||||
**Avantages :** Professionnel, historique long terme, alerting avancé
|
||||
|
||||
```bash
|
||||
# Exposer les métriques Incus
|
||||
incus config set core.metrics_address :8443
|
||||
|
||||
# Installer Prometheus + Grafana
|
||||
# (plus complexe, voir doc officielle)
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### Option 3 : Scripts + Monitoring Simple
|
||||
|
||||
Si vous préférez rester léger, surveillez manuellement avec :
|
||||
|
||||
```bash
|
||||
# Dashboard temps réel
|
||||
watch -n 2 'echo "=== RESSOURCES ===";
|
||||
free -h | head -2;
|
||||
echo "";
|
||||
echo "=== PHP-FPM ===";
|
||||
ps aux | grep "php-fpm: pool" | wc -l;
|
||||
echo "";
|
||||
echo "=== MySQL ===";
|
||||
mysql -e "SHOW STATUS LIKE \"Threads_connected\";" 2>/dev/null;
|
||||
echo "";
|
||||
echo "=== NGINX ===";
|
||||
netstat -an | grep :80 | wc -l'
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 9. Checklist de Diagnostic en Cas de Problème
|
||||
|
||||
### Coupures / Timeouts
|
||||
1. ✅ **Vérifier PHP-FPM** : `tail /var/log/php8.3-fpm.log | grep max_children`
|
||||
2. ✅ **Vérifier MySQL** : `mysql -e "SHOW PROCESSLIST;"`
|
||||
3. ✅ **Vérifier nginx** : `tail /var/log/nginx/error.log | grep -E "502|504"`
|
||||
4. ✅ **Vérifier RAM** : `free -h`
|
||||
5. ✅ **Vérifier I/O** : `iostat -x 1 5`
|
||||
|
||||
### Lenteurs
|
||||
1. ✅ **Slow queries MySQL** : `tail /var/lib/mysql/slow-query.log`
|
||||
2. ✅ **Slow PHP** : `tail /var/log/php8.3-fpm-slow.log`
|
||||
3. ✅ **CPU** : `top -bn1 | head -20`
|
||||
4. ✅ **I/O disque** : `iotop -oa`
|
||||
|
||||
### Crashes / Erreurs 502
|
||||
1. ✅ **PHP-FPM status** : `systemctl status php8.3-fpm`
|
||||
2. ✅ **OOM Killer** : `dmesg | grep -i "out of memory"`
|
||||
3. ✅ **Logs PHP** : `tail -100 /var/log/php8.3-fpm.log`
|
||||
|
||||
---
|
||||
|
||||
## 10. Optimisations Recommandées
|
||||
|
||||
### PHP-FPM
|
||||
```ini
|
||||
# /etc/php/8.3/fpm/pool.d/www.conf
|
||||
pm = dynamic
|
||||
pm.max_children = 50-100
|
||||
pm.start_servers = 10-20
|
||||
pm.min_spare_servers = 5-10
|
||||
pm.max_spare_servers = 20-35
|
||||
pm.max_requests = 500
|
||||
request_slowlog_timeout = 3s
|
||||
```
|
||||
|
||||
### MariaDB
|
||||
```ini
|
||||
# /etc/mysql/mariadb.conf.d/50-server.cnf
|
||||
max_connections = 200-500
|
||||
innodb_buffer_pool_size = 16G # 70-80% de la RAM MySQL
|
||||
slow_query_log = 1
|
||||
long_query_time = 2
|
||||
log_queries_not_using_indexes = 1
|
||||
```
|
||||
|
||||
### Nginx
|
||||
```nginx
|
||||
# /etc/nginx/nginx.conf
|
||||
worker_processes auto;
|
||||
worker_connections 2048;
|
||||
keepalive_timeout 30;
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Résumé des Seuils Critiques
|
||||
|
||||
| Métrique | Warning | Critical |
|
||||
|----------|---------|----------|
|
||||
| **RAM** | >85% | >95% |
|
||||
| **CPU Load** | >70% CPU count | >150% CPU count |
|
||||
| **Disk I/O await** | >50ms | >200ms |
|
||||
| **Disk Space** | >80% | >90% |
|
||||
| **PHP Workers** | >80% max | max_children reached |
|
||||
| **MySQL Connections** | >80% max | = max_connections |
|
||||
| **Slow Queries** | >2s | >10s ou >50/min |
|
||||
| **Nginx 502/504** | >5 en 5min | >20 en 5min |
|
||||
|
||||
---
|
||||
|
||||
## Contacts et Escalade
|
||||
|
||||
En cas d'incident critique :
|
||||
1. Vérifier les logs (`/var/log/alerts.log`)
|
||||
2. Identifier le goulot (CPU/RAM/PHP/MySQL)
|
||||
3. Appliquer le correctif approprié
|
||||
4. Documenter l'incident
|
||||
|
||||
---
|
||||
|
||||
**Date de création :** 2025-10-18
|
||||
**Dernière mise à jour :** 2025-10-18
|
||||
**Version :** 1.0
|
||||
Reference in New Issue
Block a user