As you build out your cloud resources, you may begin to feel overwhelmed with the different elements that need your attention. Even if you have smaller environments, notifications are essential in case something is about to break or has already gone wrong.
VMware vSphere includes default alarms as well as the ability to add custom alarms that can notify admins for various events, conditions, or states. You can set up and configure alarms from the vSphere web client or native application.
Alarms can be set for:
Alarms can be activated when an event occurs, when a set of conditions happen, or when a certain state is reached by an object in your vSphere inventory (storage, VM, etc).
The default vSphere alarms are fairly comprehensive, but you'll have to set the alert action for each.
Condition and state alarm monitor types can be combined and are under the same selection when setting up a new alarm. A condition trigger is a number value that must exceed a threshold, like “CPU usage greater than XX Ghz”. A state trigger refers to a predefined element from a set of states that each object might have at any given time, like “connected” or “not connected”.
Specific events are a separate monitoring option when configuring alarm definitions and cannot be combined with conditions. They refer to events within the overall environment, like “VM created” or user “Role added”.
Talk to one of our experts today.
Each inventory tab (Hosts and Clusters, Virtual Machines and Templates, Datastores, and Networks) has an Alarms tab at the top. This tab has views for Triggered Alarms and Alarm Definitions.
You can view and turn off the default alarms under Definitions. Note that these default alarms do not have an Action set until you choose one. This is also where you will add new alarms (New > Alarm).
Alarm definitions include:
Alarms can also be created through:
As an example, let’s set up an alarm that monitors the CPU usage of all virtual machines in a resource pool and sends you an e-mail when they reach a certain threshold. From the inventory, select the pool and right click, then choose Alarm > Add Alarm. Name it CPU Use and enter a description if you like. Select “Monitor for specific conditions or state”, check “Enable this alarm” then click the Triggers tab.
Under Triggers, choose Add. You’ll have many trigger types to choose from, but pick Host CPU Usage. Then set a Condition – in this case we’ll go with “Is above”. You can set the alarm for a warning level and an alert level. For this alert we’ll choose 90 as our Alert, with a Condition length of 10 minutes. This means that if the CPU use is over 90% for 10 minutes, the alarm will be triggered.
The Reporting tab is used to create a range for alarms, specifying their frequency and tolerance for a given condition. We will keep these default. Finally, move on to Actions and select Send a notification email. Enter your e-mail address under Configuration and choose how often you want an alert. Hit OK and your new alarm is configured!
Alarms are versatile and can be used in combination with the API to create notifications for almost any event or combination of factors. vSphere alarms can also be useful if your company has compliance or security mandates as they will alert you to suspicious activity like updated user permissions.