OpenStack Cluster Alerts
At Liquid Web, we are continually monitoring the functionality and health of all our servers to make sure that our customers have the best experience possible. Occasionally, we will send notifications to customers if we see resource usage that could cause problems with a server’s functionality or responsiveness. You may receive an OopenStack Cluster Alert if we see resource usage spiking above your server’s capacity for extended periods of time (30 minutes or more).
You can also use the information in your Liquid Web Account to check the resource usage on your servers.
Identifying Server Load
You can view current and historical server load information in your Liquid Web Account.
- Login to your Liquid Web Account.
- Click on Servers at the left of the screen.

- Select any of the webnodes for the project that you are investigating (all of the nodes will show very similar usage statistics due to the load balancing of the cluster).

- Scroll down to the Server Health graph located below the general information about the web node.

Calculating Server Load
Now that you’ve found the statistics about your web node activity, you can do a simple calculation to decide if your web nodes are being overloaded or underutilized.
- Determine the number of CPU cores allocated to your web node. The number can be found in the general information for the web node.
- Divide the average server load number by the total number of CPU cores for the node to get the percentage of resource usage. For example, if the server load from the graph is 2 and the webnode has 4 CPU cores, the resource utilization is 50%.

Modifying your Web Nodes
As a general rule, your OpenStack Server Cluster should be functioning between 25% and 75% resource usage over a given period of time.
Usage over 75% could affect your end users experience and may indicate the need to add a web node to your cluster to manage the resource usage.
Usage under 25% may indicate an over allocation of resources and you may want to consider removing a web node to save money until usage increases.
