Realmente, o problema eh que o keepalived retorna 1 quando é iniciado sem ser parado antes, ou seja, no segundo start consecutivo ele começa a retornar 1... Mas porque o heartbeat está iniciando os serviços novamente se eles já estão iniciados?
Obrigado 2006/5/15, Flavio Menezes Reis <[EMAIL PROTECTED]>: > Pra embassar mais... o log do que acontece quando eu desligo o > secundário. Parece que o problema está em o hearbeat novamente dar > start em tudo que está em haresources quando o secundário é desligado: > heartbeat[6047]: 2006/05/15_11:30:45 info: Received shutdown notice > from 'sempron2200_2'. > heartbeat[6047]: 2006/05/15_11:30:45 info: Resources being acquired > from sempron2200_2. > heartbeat[7102]: 2006/05/15_11:30:45 info: acquire local HA resources > (standby). > ResourceManager[7122]: 2006/05/15_11:30:45 info: Acquiring resource > group: athlon2400 IPaddr::192.168.1.50/24/eth0 drbddisk > Filesystem::/dev/drbd0::/mnt/data postgresql-8.1 cluster0 keepalived > IPaddr[7147]: 2006/05/15_11:30:45 INFO: IPaddr Running OK > heartbeat[7103]: 2006/05/15_11:30:45 info: Local Resource acquisition > completed. > IPaddr[7166]: 2006/05/15_11:30:45 INFO: IPaddr Running OK > ResourceManager[7122]: 2006/05/15_11:30:45 info: Running > /etc/ha.d/resource.d/drbddisk start > Filesystem[7485]: 2006/05/15_11:30:45 INFO: /mnt/data is mounted > (running) > Filesystem[7421]: 2006/05/15_11:30:45 INFO: Filesystem Running OK > ResourceManager[7122]: 2006/05/15_11:30:45 info: Running > /etc/init.d/postgresql-8.1 start > ResourceManager[7122]: 2006/05/15_11:30:46 info: Running > /etc/ha.d/resource.d/cluster0 start > ResourceManager[7122]: 2006/05/15_11:30:47 info: Running > /etc/init.d/keepalived start > ResourceManager[7122]: 2006/05/15_11:30:47 ERROR: Return code 1 from > /etc/init.d/keepalived > ResourceManager[7122]: 2006/05/15_11:30:47 CRIT: Giving up resources > due to failure of keepalived > ResourceManager[7122]: 2006/05/15_11:30:47 info: Releasing resource > group: athlon2400 IPaddr::192.168.1.50/24/eth0 drbddisk > Filesystem::/dev/drbd0::/mnt/data postgresql-8.1 cluster0 keepalived > ResourceManager[7122]: 2006/05/15_11:30:47 info: Running > /etc/init.d/keepalived stop > ResourceManager[7122]: 2006/05/15_11:30:47 info: Running > /etc/ha.d/resource.d/cluster0 stop > ResourceManager[7122]: 2006/05/15_11:30:49 info: Running > /etc/init.d/postgresql-8.1 stop > ResourceManager[7122]: 2006/05/15_11:30:51 info: Running > /etc/ha.d/resource.d/Filesystem /dev/drbd0 /mnt/data stop > Filesystem[7755]: 2006/05/15_11:30:52 INFO: Filesystem Success > ResourceManager[7122]: 2006/05/15_11:30:52 info: Running > /etc/ha.d/resource.d/drbddisk stop > ResourceManager[7122]: 2006/05/15_11:30:53 info: Running > /etc/ha.d/resource.d/IPaddr 192.168.1.50/24/eth0 stop > IPaddr[7988]: 2006/05/15_11:30:54 INFO: /sbin/route -n del -host > 192.168.1.50 > IPaddr[7988]: 2006/05/15_11:30:54 INFO: /sbin/ifconfig eth0:0 > 192.168.1.50 down > IPaddr[7988]: 2006/05/15_11:30:54 INFO: IP Address 192.168.1.50 released > IPaddr[7883]: 2006/05/15_11:30:54 INFO: IPaddr Success > heartbeat[7102]: 2006/05/15_11:30:54 info: local HA resource > acquisition completed (standby). > heartbeat[6047]: 2006/05/15_11:30:54 info: Standby resource > acquisition done [foreign]. > harc[8041]: 2006/05/15_11:30:54 info: Running /etc/ha.d/rc.d/status status > mach_down[8050]: 2006/05/15_11:30:54 info: > /usr/lib/heartbeat/mach_down: nice_failback: foreign resources > acquired > mach_down[8050]: 2006/05/15_11:30:54 info: mach_down takeover > complete for node sempron2200_2. > heartbeat[6047]: 2006/05/15_11:30:54 info: mach_down takeover complete. > > > 2006/5/15, Flavio Menezes Reis <[EMAIL PROTECTED]>: > > Olá! > > > > Tava quase tudo pronto aqui no cluster, com auto failback on e tudo, o > > problema é que não havia experimentado desligar o secundário e > > incrivelmente, quando faço isto, o keepalived vem com um Return Code = > > 1 (RC=1) que faz com que o heartbeat comece a dar stop em todos os > > script do arquivo haresource o que acaba por derrubar o servidor... > > Porque o keepalived retorna este RC=1??? > > > > Obrigado a todos. Daqui a pouco vou remeter um relato completo de como > > construi este cluster para trocar experiências com vocês. > > > > Abraços > > > > -- > > Flávio Menezes dos Reis > > Bacharelando em Sistemas de Informação - Ulbra - Torres - RS > > [EMAIL PROTECTED] > > > > > -- > Flávio Menezes dos Reis > Bacharelando em Sistemas de Informação - Ulbra - Torres - RS > [EMAIL PROTECTED] > -- Flávio Menezes dos Reis Bacharelando em Sistemas de Informação - Ulbra - Torres - RS [EMAIL PROTECTED] _______________________________________________ Linux-HA mailing list [email protected] http://listas.linuxchix.org.br/mailman/listinfo/linux-ha
