Refresh Nodes with Updated Configuration Data

How to implement Cray service changes when the system is not being rebooted.

This procedure assumes that configuration data has been changed.

Whenever data in a global or CLE config set has been changed, it is necessary to update and validate that config set and run cray-ansible on the SMW (for global config set changes) and any affected CLE nodes (for CLE config set changes) in order to apply the configuration changes. If the system will be rebooted, the steps to run cray-ansible are not needed because cray-ansible is run automatically when the system boots.

This procedure updates the CLE config set, if needed, validates it, then runs cray-ansible to refresh the config set data stored on the CLE nodes. Using cfgset update ensures that all pre- and post-configuration scripts get run. Running cray-ansible on a CLE node triggers a refresh of the CLE config set cache on that node and applies configuration changes on the node. If the node will be rebooted, the step to run cray-ansible is not needed because cray-ansible is run automatically when the system boots.

(If cfgset update not already run) Update the config set.
```
smw# cfgset update p0
```
Validate the config set.
```
smw# cfgset validate p0
```
(If CLE nodes will not be rebooted) Run Ansible plays on the CLE nodes.
After the CLE config set has been updated, refresh the local config set cache to pull any config set changes to the node and run cray-ansible to apply them on the node.
```
hostname# /opt/cray/imps-distribution/default/bin/refresh.py
hostname# /etc/init.d/cray-ansible start
```