AdminSysGNULinux

Outils pour utilisateurs

Outils du site

Piste :
tech:cluster_pacemaker_et_corosync

Différences

Ci-dessous, les différences entre deux révisions de la page.

Lien vers cette vue comparative

Les deux révisions précédentesRévision précédente
Prochaine révision		Révision précédente
tech:cluster_pacemaker_et_corosync [2025/03/29 21:03] – Jean-Baptistetech:cluster_pacemaker_et_corosync [2026/06/07 19:12] (Version actuelle) – modification externe 127.0.0.1
Ligne 1: Ligne 1:
 +<!DOCTYPE markdown>
 +{{tag>Brouillon Cluster}}
 +
 +# Cluster Pacemaker et Corosync
 +
 +Voir :
 +* [[cluster_script_de_test_stonith_fencing]]
 +* http://clusterlabs.org/doc/en-US/Pacemaker/1.0/html/Pacemaker_Explained/s-intro-architecture.html
 +* https://wiki.debian.org/Debian-HA/ClustersFromScratch
 +* http://blogduyax.madyanne.fr/haute-disponibilite-avec-corosync-et-pacemaker.html
 +* https://www.sebastien-han.fr/blog/2011/07/04/introduction-au-cluster-sous-linux/
 +* https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/html/High_Availability_Add-On_Administration/ch-startup-HAAA.html
 +* Paquet resource-agents
 +
 +Pcs Pacemaker Configuration System :
 +* https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/html/High_Availability_Add-On_Reference/s1-clusternodemanage-HAAR.html
 +* https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/pdf/High_Availability_Add-On_Administration/Red_Hat_Enterprise_Linux-7-High_Availability_Add-On_Administration-en-US.pdf
 +* https://www.digitalocean.com/community/tutorials/how-to-create-a-high-availability-setup-with-pacemaker-corosync-and-floating-ips-on-centos-7
 +* https://skcave.wordpress.com/2014/11/04/creating-high-availability-cluster-with-centos-7/
 +* https://www.digitalocean.com/community/tutorials/how-to-set-up-an-apache-active-passive-cluster-using-pacemaker-on-centos-7
 +* http://bigthinkingapplied.com/creating-a-linux-cluster-in-red-hatcentos-7/
 +
 +
 +## Conf install
 +
 +### Sur tous les nœuds
 +
 +`/etc/hosts`
 +~~~
 +192.168.56.21  node2
 +192.168.56.22  node1
 +~~~
 +
 +~~~bash
 +yum install -y corosync pcs #pacemaker
 +~~~
 +
 +
 +
 +
 +
 +## PCS Pacemaker Configuration System
 +
 +~~~bash
 +yum install corosync pcs #pacemaker
 +
 +
 +
 +#passwd hacluster
 +
 +echo "passwd" | passwd hacluster --stdin
 +
 +systemctl start pcsd.service
 +systemctl enable pcsd.service
 +
 +#unset http_proxy
 +export NO_PROXY=localhost,127.0.0.1,node1,node2
 +pcs cluster auth node1 node2
 +
 +pcs cluster setup --start --name my_cluster node1 node2
 +#pcs cluster setup --name my_cluster node1 node2
 +
 +pcs cluster start node1
 +#pcs cluster auth
 +
 +pcs cluster start --all
 +pcs cluster enable --all
 +
 +
 +pcs resource create myvip IPaddr2 ip=192.168.56.100
 +pcs status resources
 +
 +pcs property set stonith-enabled=false
 +# pcs property set no-quorum-policy=ignore
 +
 +# http://clusterlabs.org/doc/en-US/Pacemaker/1.1/html/Clusters_from_Scratch/_prevent_resources_from_moving_after_recovery.html
 +pcs resource defaults resource-stickiness=100
 +
 +# pcs resource
 +# pcs resource defaults
 +~~~
 +
 +~~~bash
 +pcs status
 +pcs resource --full
 +pcs resource
 +~~~
 +
 +`/etc/corosync/corosync.conf`
 +~~~c
 +totem {
 +    version: 2
 +    secauth: off
 +    cluster_name: my_cluster
 +    transport: udpu
 +}
 +
 +nodelist {
 +    node {
 +        ring0_addr: node1
 +        nodeid: 1
 +    }
 +
 +    node {
 +        ring0_addr: node2
 +        nodeid: 2
 +    }
 +}
 +
 +quorum {
 +    provider: corosync_votequorum
 +    two_node: 1
 +}
 +
 +logging {
 +    to_logfile: yes
 +    logfile: /var/log/cluster/corosync.log
 +    to_syslog: yes
 +}
 +~~~
 +
 +Autre exemple de fichier de conf
 +
 +`/etc/corosync/corosync.conf`
 +~~~c
 +totem {
 + version: 2
 + secauth: off
 + cluster_name: mycluster-1
 + transport: udpu
 + token: 5000
 +}
 +
 +nodelist {
 +  node {
 +        ring0_addr: mycluster-a1
 +        nodeid: 1
 +       }
 +  node {
 +        ring0_addr: mycluster-b1
 +        nodeid: 2
 +       }
 +}
 +
 +quorum {
 + provider: corosync_votequorum
 + auto_tie_breaker: 0
 + two_node: 1
 +}
 +
 +logging {
 + to_syslog: yes
 +}
 +
 +~~~
 +
 +
 +### Sécurité
 +
 +#### Sur nœud 1
 +
 +~~~bash
 +corosync-keygen -l
 +~~~
 +
 +Copie de la clef sur le nœud 2
 +~~~bash
 +scp /etc/corosync/authkey node2:/etc/corosync/authkey
 +~~~
 +
 +~~~bash
 +sed -i -e 's/secauth: off/secauth: on/' /etc/corosync/corosync.conf
 +pcs cluster sync
 +~~~
 +
 +
 +### Bascule 
 +
 +~~~bash
 +pcs status resources
 +~~~
 +
 +~~~
 + myvip  (ocf::heartbeat:IPaddr2):       Started node1
 +~~~
 + 
 +~~~bash
 +pcs cluster standby node1
 +~~~
 +
 +~~~bash
 +pcs status resources
 +~~~
 +
 +~~~
 + myvip  (ocf::heartbeat:IPaddr2):       Started node2
 +~~~
 +
 +Réactivation node1
 +~~~bash
 +pcs cluster unstandby node1
 +~~~
 +
 +Rebascule de la VIP **myvip** sur node1
 +~~~bash
 +pcs resource move myvip node1
 +~~~
 +
 +Mise à zéro compteur
 +~~~bash
 +pcs resource relocate clear
 +~~~
 +
 +
 +### Vérif config
 +
 +~~~bash
 +crm_verify -LV
 +
 +crm_verify -LVVV
 +~~~
 +
 +
 +## Autres
 +
 +Logs
 +~~~bash
 +tailf /var/log/cluster/corosync.log
 +~~~
 +
 +Interface Web
 +* https://192.168.56.21:2224
 +* https://192.168.56.22:2224
 +
 +Compte **hacluster**
 +
 +
 +## Pb 
 +
 +### Connection to cluster failed: Transport endpoint is not connected
 +
 +** Pas nécessaire / pas de pb avec pcs**
 +
 +~~~bash
 +crm_mon -1
 +~~~
 +
 +~~~
 +Connection to cluster failed: Transport endpoint is not connected
 +~~~
 +
 +~~~bash
 +systemctl restart pacemaker
 +~~~
 +
 +~~~bash
 +crm_mon -1
 +~~~
 +
 +~~~
 +Last updated: Fri Oct  7 11:59:55 2016          Last change: Fri Oct  7 11:03:51 2016
 +Stack: unknown
 +Current DC: NONE
 +0 nodes and 0 resources configured
 +~~~
 +
 +
 +Pour ne plus à devoir démarrer manuelelement **pacemaker** après **corosync**
 +
 +`/etc/corosync/corosync.conf`
 +~~~
 +service {
 +        # Load the Pacemaker Cluster Resource Manager
 +        name: pacemaker
 +        ver:  1
 +}
 +~~~
 +
 +
 +### Effacer les "Failed Actions"
 +
 +~~~
 +pcs status
 +Cluster name: my_cluster
 +Last updated: Fri Oct 14 16:40:14 2016          Last change: Fri Oct 14 16:36:04 2016 by root via cibadmin on node1
 +Stack: corosync
 +Current DC: node2 (version 1.1.13-10.el7_2.4-44eb2dd) - partition with quorum
 +2 nodes and 2 resources configured
 +
 +Node node1: UNCLEAN (online)
 +Online: [ node2 ]
 +
 +Full list of resources:
 +
 + myvip  (ocf::heartbeat:IPaddr2):       Started node1
 + plopa  (stonith:fence_ssh):    Started node2
 + clusterfs      (ocf::heartbeat:Filesystem):     ORPHANED FAILED node1
 +
 +Failed Actions:
 +* clusterfs_stop_0 on node1 'unknown error' (1): call=38, status=Timed Out, exitreason='none',
 +    last-rc-change='Fri Oct 14 14:43:54 2016', queued=0ms, exec=60003ms
 +
 +
 +PCSD Status:
 +  node1: Online
 +  node2: Online
 +
 +Daemon Status:
 +  corosync: active/enabled
 +  pacemaker: deactivating/enabled
 +  pcsd: active/enabled
 +
 +~~~
 +
 +Source : https://supportex.net/blog/2011/09/cleaning-failed-actions-pacemakercorosync-cluster-setup/
 +
 +Solution
 +~~~bash
 +#crm_resource -P
 +pcs resource cleanup
 +~~~
 +
 +
 +### Pb node2: Unable to connect to node2 ([Errno 111] Connection refused)
 +
 +~~~
 +journalctl -u pacemaker.service -f -p warning
 +-- Logs begin at ven. 2017-06-09 18:23:40 CEST. --
 +juin 12 11:28:23 pc-2 attrd[170467]:    error: Cluster connection failed
 +juin 12 11:28:23 pc-2 pacemakerd[170463]:    error: Managed process 170467 (attrd) dumped core
 +juin 12 11:28:23 pc-2 pacemakerd[170463]:    error: The attrd process (170467) terminated with signal 11 (core=1)
 +juin 12 11:28:23 pc-2 cib[170464]:    error: Could not connect to the Cluster Process Group API: 11
 +juin 12 11:28:23 pc-2 cib[170464]:     crit: Cannot sign in to the cluster... terminating
 +juin 12 11:28:23 pc-2 pacemakerd[170463]:  warning: The cib process (170464) can no longer be respawned, shutting the cluster down.
 +juin 12 11:28:23 pc-2 crmd[170469]:  warning: Couldn't complete CIB registration 1 times... pause and retry
 +juin 12 11:28:23 pc-2 crmd[170469]:  warning: FSA: Input I_SHUTDOWN from crm_shutdown() received in state S_STARTING
 +juin 12 11:28:23 pc-2 pacemakerd[170463]:    error: The attrd process (170470) terminated with signal 15 (core=0)
 +juin 12 11:28:33 pc-2 stonith-ng[170465]:    error: Could not connect to the CIB service: Transport endpoint is not connected (-107)
 +
 +Sur les deux noeuds
 +pcs cluster stop --all
 +node2: Unable to connect to node2 ([Errno 111] Connection refused)
 +
 +lsof -Pn -i TCP:2224
 +
 +pacemakerd[170758]:    error: The attrd process (170765) terminated with signal 15 (core=0)
 +
 +systemctl status corosync
 +juin 12 12:39:02 pc-1 corosync[116364]:  [MAIN  ] Denied connection attempt from 189:189
 +juin 12 12:39:02 pc-1 corosync[116364]:  [QB    ] Invalid IPC credentials (116364-116387-2).
 +juin 12 12:39:02 pc-1 corosync[116364]:  [MAIN  ] Denied connection attempt from 189:189
 +juin 12 12:39:02 pc-1 corosync[116364]:  [QB    ] Invalid IPC credentials (116364-116384-2).
 +
 +
 +ps -C cib,stonithd,pcsd,lrmd,attrd,pengine,crmd
 +# Only pcsd
 +
 +ss -lx | grep @cib_
 +# No unix socket open
 +~~~
 +
 +
  
Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki