Error Message

The /var/log/messages file reports:

AgentFramework[6019]: [ID 702911 daemon.notice] VCS ERROR V-16-20028-1005 CVMVxconfigd:cvm_vxconfigd:monitor:vxconfigd core dumped. Moving the core to core.19611. It is probable but not necessary that this core dump caused vxconfigd to die. You must match timestamps of this message and the vxconfigd core to be sure.

The core dump happens because of duplicate records, as seen in the output below from vxdisk -oalldgs list with the error lines:

purestorage3_000113f0 auto:cdsdisk    -            (DHWARE_ORACLE) error
purestorage3_000113f0 auto:cdsdisk    purestorage3_000113f0  DHWARE_ORACLE online thinrclm

purestorage3_000113f1 auto:cdsdisk    -            (DHWARE_ORACLE) error
purestorage3_000113f1 auto:cdsdisk    purestorage3_000113f1  DHWARE_ORACLE online thinrclm
 

First, run hastop -all to stop VCS and drop Cluster Volume Manager (CVM) membership across the cluster.

For each node of the cluster:

1) Rename /etc/vx/salrecs to /etc/vx/salrecs.bk and /etc/vx/disk.info to /etc/vx/disk.info.bk
2) Run vxconfigd -kr reset
3) Check the vxdisk list output and confirm there are no duplicate records.
4) Join the nodes back to the cluster using hastart.
5) Check the vxdisk list output once again and confirm there are no duplicate records.