Pages : 1
#1 Le 29/06/2016, à 13:58
- chris92
Pacemaker/Corosync sur Ubuntu 16.04
Bonjour
J'ai une question concernant un cluster Pacemaker que j'essaie de mettre en place. Pacemaker (qui gère les ressources ) utilise corosync comme layer de communication entre mes noeuds.
Mon problème : après 45 secondes, mes noeuds arrêtent le process Pacemaker (le process corosync restant exécuté).
Voici donc un extrait de mon fichier de logs /var/log/pacemaker.log (l'arret du process se passe à 14:55:36)
Jun 29 14:55:01 [30193] bos3mep stonith-ng: info: qb_ipcs_us_publish: server name: stonith-ng
Jun 29 14:55:01 [30193] bos3mep stonith-ng: info: pcmk_cpg_membership: Node 1 joined group stonith-ng (counter=0.0)
Jun 29 14:55:01 [30193] bos3mep stonith-ng: info: pcmk_cpg_membership: Node 1 still member of group stonith-ng (peer=node1, counter=0.0)
Jun 29 14:55:36 [30195] bos3mep attrd: info: qb_ipcs_us_withdraw: withdrawing server sockets
Jun 29 14:55:36 [30195] bos3mep attrd: info: crm_xml_cleanup: Cleaning up memory from libxml2
Jun 29 14:55:36 [30197] bos3mep crmd: notice: crm_signal_dispatch: Invoking handler for signal 15: Terminated
Jun 29 14:55:36 [30197] bos3mep crmd: warning: do_log: FSA: Input I_SHUTDOWN from crm_shutdown() received in state S_STARTING
Jun 29 14:55:36 [30197] bos3mep crmd: notice: do_state_transition: State transition S_STARTING -> S_STOPPING [ input=I_SHUTDOWN cause=C_SHUTDOWN origin=crm_shutdown ]
Jun 29 14:55:36 [30197] bos3mep crmd: info: crm_cluster_disconnect: Disconnecting from cluster infrastructure: corosync
Jun 29 14:55:36 [30197] bos3mep crmd: notice: terminate_cs_connection: Disconnecting from Corosync
Jun 29 14:55:36 [30197] bos3mep crmd: info: cluster_disconnect_cpg: No CPG connection
Jun 29 14:55:36 [30197] bos3mep crmd: info: terminate_cs_connection: No Quorum connection
Jun 29 14:55:36 [30197] bos3mep crmd: info: crm_cluster_disconnect: Disconnected from corosync
Jun 29 14:55:36 [30197] bos3mep crmd: info: do_exit: Performing A_EXIT_0 - gracefully exiting the CRMd
Jun 29 14:55:36 [30197] bos3mep crmd: info: crm_xml_cleanup: Cleaning up memory from libxml2
Jun 29 14:55:36 [30196] bos3mep pengine: notice: crm_signal_dispatch: Invoking handler for signal 15: Terminated
Jun 29 14:55:36 [30196] bos3mep pengine: info: qb_ipcs_us_withdraw: withdrawing server sockets
Jun 29 14:55:36 [30196] bos3mep pengine: info: crm_xml_cleanup: Cleaning up memory from libxml2
Jun 29 14:55:36 [30194] bos3mep lrmd: notice: crm_signal_dispatch: Invoking handler for signal 15: Terminated
Jun 29 14:55:36 [30194] bos3mep lrmd: info: qb_ipcs_us_withdraw: withdrawing server sockets
Jun 29 14:55:36 [30194] bos3mep lrmd: info: crm_xml_cleanup: Cleaning up memory from libxml2
Jun 29 14:55:36 [30193] bos3mep stonith-ng: notice: crm_signal_dispatch: Invoking handler for signal 15: Terminated
Jun 29 14:55:36 [30193] bos3mep stonith-ng: info: qb_ipcs_us_withdraw: withdrawing server sockets
Jun 29 14:55:36 [30193] bos3mep stonith-ng: info: crm_xml_cleanup: Cleaning up memory from libxml2
Jun 29 14:55:36 [30165] bos3mep pacemakerd: info: crm_xml_cleanup: Cleaning up memory from libxml2
Mon environnement :
pacemaker 1.1.14 sur Ubuntu 16.04
corosync 2.3.5
Ma configuration :
Les clés ont été partagées entre les 2 noeuds de mon cluster
Voici le fichier /etc/corosync/cosorync.conf
totem {
version: 2
cluster_name: clusterTest
secauth: off
transport:udpu
interface {
ringnumber: 0
bindnetaddr: 10.10.100.0
broadcast: yes
mcastport: 5405
}
}
nodelist {
node {
ring0_addr: 10.10.100.157
name: node1
nodeid: 1
}
node {
ring0_addr: 10.10.100.158
name: node2
nodeid: 2
}
}
Merci d'avance les gens
Modération : merci à l'avenir d'utiliser les balises code (explications ici).
Dernière modification par cqfd93 (Le 29/06/2016, à 15:49)
Hors ligne
Pages : 1