There are several possible causes and solutions when you command a node online, but it does not go online or turns itself offline shortly after going online.
For more details about the product and how to configure features, click Help or
press F1.
Steps
-
If you have just updated the engines or you are using an evaluation license, open the Administration Configuration view and browse to Licenses to check that your licenses are valid for the version of engines you are using.
-
If the nodes are in a cluster, and only one node at a time stays online, try the following troubleshooting steps:
- Check whether the cluster is in Standby mode. Standby mode keeps one node online at a time and uses the other nodes as backups in case the online node fails.
- Refresh the policy of the cluster and check that the installation is successful so that no nodes roll back to the previous configuration. All nodes in the cluster must have the same configuration that has been installed in the same policy installation operation. You might have to adjust the rollback timeout in the cluster’s properties if policy rollback on some node is the problem.
- Check for alerts in the Logs view about tests failing. Check if any of the failed tests are configured to turn the node offline when they fail. The tester leaves one node in a cluster online even if the test fails on all nodes. If you see a test failure, it might indicate a genuine problem that you need to solve or the test might be misconfigured and might have to be disabled or reconfigured.
-
See the Logs view for any alerts or logs regarding the functioning of the nodes.
Certain internal error conditions (for example, heartbeat connection failures or missing certificates) can cause nodes to go offline. These events are shown as logs and alerts.