Déroulement de l'atelier "Surveiller et améliorer les performances de son serveur"

Plan

1. Énumération de quelques outils
2. Quelques exemples d'utilisation à Montréal (nagios, munin, logcheck) et à Dakar (mon)
3. Échanges sur les expériences des participants
4. Recommandations en matière de supervision
5. Comment améliorer les performances de son serveur ?

Déroulement

1- Énumération de quelques outils

• le faire soi-même (pas productif)
  • Suivi des journaux d’évènements (les logs)
    • auth.log
    • syslog
    • mail.log
    • dmesg
  • quelques commandes
    • top- Process Activity Command
    • vmstat – System Activity, Hardware and System Information
    • w – Find Out Who Is Logged on And What They Are Doing
    • ps – Displays The Processes
    • uptime – Tell How Long The System Has Been Running
    • free – Memory Usage
    • iptraf – Real-time Network Statistics
    • iostat – Average CPU Load, Disk Activity
    • netstat – Network Statistics
    • ss – Network Statistics
    • tcpdump – Detailed Network Traffic Analysis
    • mpstat – Multiprocessor Usage
    • sar – Collect and Report System Activity
    • pmap – Process Memory Usage
    • ipfm - a bandwidth analysis
• supervision active (envoi d'alertes)
  • Nagios
  • Logcheck / logwatch
• supervision passive
  • Cacti
  • Munin
  • Mrtg
  • Zabbix

  En deux liens : http://www.debianhelp.co.uk/monitortools.htm et http://www.cyberciti.biz/tips/top-linux-monitoring-tools.html

2- Exemples de Montréal

1. Nagios

   • ***** Nagios  *****
     
     Notification Type: PROBLEM
     
     Service: Espace disque
     Host: vz-aufhorsite
     State: WARNING for 0d 0h 3m 1s
     Address: 204.136.13.20
     
     Info:
     
     WARNINGbr/: 90% 2386/2818MB used (warning=90% critical=95%)br
     
     Date/Time: Thu Aug 4 13:40:40 EDT 2011

     ***** Nagios  *****
     
     Notification Type: PROBLEM
     
     Service: Espace disque
     Host: vz-aufhorsite
     State: CRITICAL for 0d 0h 0m 1s
     Address: 204.136.13.20
     
     Info:
     
     CRITICALbr/: 95% 2533/2818MB used (warning=90% critical=95%)br
     
     Date/Time: Fri Aug 5 17:00:40 EDT 2011 

     ***** Nagios  *****
     
     Notification Type: RECOVERY
     
     Service: Espace disque
     Host: vz-aufhorsite
     State: OK for 0d 0h 0m 1s
     Address: 204.136.13.20
     
     Info:
     
     OK
     
     Date/Time: Sun Aug 7 06:25:40 EDT 2011

     **** Nagios *****
     
     Notification Type: PROBLEM
     Host: nfs-mtl
     State: DOWN for 0d 0h 0m 0s
     Address: 10.36.1.200
     Info:
     
     CRITICAL - Host Unreachable (10.36.1.200)
     
     Date/Time: Mon Jul 18 17:18:00 EDT 2011

2. logcheck

   • ecurity Events for su
     =-=-=-=-=-=-=-=-=-=-=-
     Jul 28 15:03:23 10.36.0.17 su[3079]: pam_unix(su:auth): authentication failure; logname=xxxxx uid=1008 euid=0 tty=pts/1 ruser=xxxxx rhost=  user=ftp
     Jul 28 15:03:25 10.36.0.17 su[3079]: FAILED su for ftp by xxxxx
     
     Security Events for sudo
     =-=-=-=-=-=-=-=-=-=-=-=-
     Jul 28 15:01:48 10.36.0.17 sudo: pam_unix(sudo:auth): authentication failure; logname=xxxxx uid=0 euid=0 tty=/dev/pts/1 ruser= rhost=  user=xxxxx

3. Munin

   • suivi des onduleurs : http://superca-munin.ca.auf/onduleurs.html

   • suivi de ressources systèmes, exemple de la mémoire :