Manage cluster nodes
You can check whether cluster nodes are running from the Nodes page.
You can access the Nodes page from the menu icon: ( ).
For each node in the cluster, you can see:
- The IP address of the node
- Whether the node is running or down
- How many CPU cores the node is using
- The GB of memory the node is using
- The GB of storage (disk) the node is using
Managing GlusterFS logs
The GlusterFS logs are by default stored under
/var/log/glusterfs on each of the
master nodes of the cluster. These logs can grow and should be periodically deleted or archived.
You can check the size of the logs by running
du -sh /var/log/glusterfs/.
You can remove the logs by running
rm -f /var/log/glusterfs/*.gz
GlusterFS log rotation is managed by the Linux
logrotate.d process, which by
default retains GlusterFS logs for 52 days. By modifying the
rotate value in
/etc/logrotate.d/glusterfs, you can reduce the number of days for logs that are
retained by the
Tip: Do you need more detailed information?
If you click the node name, you can see the CPU, storage, and memory consumption for the node over the last 20 minutes.
If you expand the node name, you can see the operating system partition and disk partition where the node is running. You can see the size of the partition and how much of the partition is in use.
Kubernetes is designed to handle node outages, so you can still use Watson Studio Local even if some of your nodes are down. However, if a node is down, you must attempt to diagnose and resolve the problem to avoid outages if other nodes fail.
Contact IBM Software Support if you cannot determine why the node failed or how to restart the node.
You can add additional compute or deployment nodes by clicking Add node. See System requirements for Watson Studio Local for the node requirements.