Skip to content

Latest commit

 

History

History
36 lines (29 loc) · 1.06 KB

aix_hacmp.md

File metadata and controls

36 lines (29 loc) · 1.06 KB
Raw

HACMP

View cluster state:

/usr/es/sbin/cluster/clstat -o

cd /usr/es/sbin/cluster/utilities
./clRGinfo
./cltopinfo -i   # heartbeat info
./cltopinfo -m   # check no heartbeats missed

To stop cluster:

  • smitty clstop, select all nodes in cluster and Bring Resource Groups Offline.
  • tail -f /var/hacmp/log/hacmp.out

PowerHA 7.1

  • cluster repository disk (1GB)
  • stretched cluster would have 1 repo disk
  • cluster survives if repo disk fails, but topo commands will failed until fixed
  • can define a backup repo disk, but is a manual backup
  • CAA cluster always active
  • clshowsrv -v shows processes; clcomd name has changed
  • source ip means packet response from alias address is correct
  • netstat -i shows service alias first
  • clcmd distributes commands to cluster nodes
  • linked cluster would have repo disk disk in each remote site

HyperSwap

  • in enterprise edition
  • nodes need to see both storage servers

split/merge

  • after split site, one is prod
  • on merge, who becomes prod ? An option is available to request manual intervention.