[ClusterLabs] Unknown error on an interface
del Cura Ferreira, Francisco
externo.fcf at maec.es
Mon Feb 20 09:05:02 UTC 2017
First of all, I'm a begineer in this field but I got to manage Pacemaker on a Ubuntu 10.04 cluster fw. That's is, 2 fw in a Active/Passive environment.
Last Thursday I changed the netmask of an interface (from /28 to /27) on both appliances (but I didn't change the netmask in the crm configure section). Since then, the cluster was down for 20 minutes (no internet, no mail, everything down). Don't know how but after that 20 minutes, the node is always passive (Node2) got the Active rol and so far it keeps it.
Right now, in Node1 I can see the next (through the crm_mon -fo1 command):
IP_Cluster_vlan493: migration-threshold=1000000 fail-count=1000000
+ (849) start: rc=1 (unknown error)
+ (850) stop: rc=0 (ok)
Node 2 has got the same configuration in /etc/network/interfaces and crm than Node1 but there is no problem with that interface:
+ (415) stop: rc=0 (ok)
+ (437) start: rc=0 (ok)
+ (438) monitor: interval=10000ms rc=0 (ok)
The crm_verify -VL you can see:
Node1# crm_verify -VL
crm_verify: 2017/02/20_09:29:32 WARN: unpack_rsc_op: Processing failed op IP_Cluster_vlan493_start_0 on TAfwx01: unknown error (1)
crm_verify: 2017/02/20_09:29:32 WARN: common_apply_stickiness: Forcing IP_Cluster_vlan493 away from TAfwx01 after 1000000 failures (max=1000000)
The output of that command in Node2 is:
Node2# crm_verify -VL
crm_verify: 2017/02/20_09:56:32 WARN: unpack_rsc_op: Processing failed op IP_Cluster_vlan493_start_0 on Node1: unknown error (1)
crm_verify: 2017/02/20_09:56:33 WARN: common_apply_stickiness: Forcing IP_Cluster_vlan493 away from Node1 after 1000000 failures (max=1000000)
If both of them's got the same configuration, how is it possible vlan493 is working fine in Node2 but fails in Node1??
Thanks a lot for your help
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Users