Restart After Planned Shutdown
Software AG recommends that you always back up your database before performing any changes to the cluster, even if you do not plan to restore the backup data afterwards.
By default, the node you start first becomes the primary node of the cluster. The primary node is usually the node that is last stopped before a graceful shutdown.
To identify which node was last stopped and has the highest sequence number (seqno), compare either the value of seqno of each node in $ADADATADIR/dbxxx/grastate.dat or execute the following command: adanuc dbid=dbid noplog options=autorestart_only.
The node with the largest seqno value is the node last stopped and should contain safe_to_bootstrap: 1 in the grastate.dat file.
To restart a cluster after a planned shutdown:
1. Ensure grastate.dat contains safe_to_bootstrap: 1 and CLUSTER_URL is not set for the node that you plan to start first.
2. Set the CLUSTER_BOOTSTRAP environment variable for the node with the highest seqno using the following command:
export CLUSTER_BOOTSTRAP=
3. Start the primary node using adastart dbid.
4. Remove the CLUSTER_BOOTSTRAP environment variable using the following command:
unset CLUSTER_BOOTSTRAP
5. For the secondary node(s), ensure that the CLUSTER_URL setting in the DBdbid.INI file is valid.
CLUSTER_BOOTSTRAP must not be set.
6. Start the secondary node(s) using adastart dbid.
If CLUSTER_BOOTSTRAP is set (via the export CLUSTER_BOOTSTRAP= command or as part of the NUCPARMS section in DBdbid.INI), the CLUSTER_URL setting will be ignored during startup.