Viewing the Cluster Status

Katana has a status page which shows the utilization of the cluster as well as individual nodes. It also allows you to look at cluster and node utilization over different time frames. As well as providing and graphing load information the cluster status page allows you to get detailed information on precisely what is happening with all aspects of the network, CPU, local disk and memory either on a cluster wide or a blade by blade basis.

This information can be combined with the information in the list of Katana nodes to allow you to check the current status and availability of nodes that have been purchased by your school or research group.

If you want more information then you can type pestat to list all the nodes of the cluster along with what jobs are running on those nodes, node memory usage, node load and if a node is able to accept more jobs. To look at the status of a particular node then you can use the grep command to filter results.

For example to see what jobs are running on Katana node kc08b03 (purchased by the Faculty of Science) type

pestat | grep kc08b03

or if you wish to see the column headings

pestat | grep -E 'node|kc08b03'

You can then use the job monitoring commands to get detailed information on how long specific jobs have to run and when a particular node may be available for use.