You can schedule nodes to be automatically removed from the QueryGrid Manager cluster that have been offline for a specified amount of time.
Only enable this feature on clusters that frequently scale without QueryGrid automatically detecting the nodes that are removed, such as with a Hadoop platform.
The following instructions are specific to the command line, but the feature can also be enabled when adding or editing a system in the Viewpoint QueryGrid portlet.
- Log on to the system where QueryGrid Manager is running.
- From /opt/teradata/tdqgm/bin, run the auto-node-delete.sh script to enable or disable this feature.Valid values are 1-1000 hours for specifying offline time. If a node is pending deletion, disabling this feature prevents the node from being deleted.
[root@system1 bin]# ./auto-node-delete.sh -h Starting auto-node-delete command, just a moment... usage: auto-node-delete [-d] [-e] [-h] [-m <host>] [-n] [-s <system>] [-t <time>] Configure automatic node deletion for a system. options: -d,--disable Flag to disable the node auto-delete feature. -e,--enable Flag to enable the node auto-delete feature. -h,--help Print this message -m,--manager <host> The hostname of the QueryGrid manager to connect to. Defaults to localhost. -n,--no-ssl Use HTTP instead of HTTPS to connect to the manager. -s,--system <system> The name of the system. -t,--time <time> Number of hours a node is offline before being auto deleted when the feature is enabled.