Monitoring and Scaling VMs

Basic control and scaling concepts

Concept	Description
Metric	A built-in metric that is provided by a hypervisor or provider (e.g. CPU usage) or a custom metric that the user creates and populates using API calls
Alarm	An alarm activates when a metric passes a certain threshold. If you imagine a dashboard for your metrics, alarms are like red lights that light up when conditions change, for example, when there is a problem. See Manage cloud alarms and Infrastructure Alarms
Alert	An alert enables you to configure notifications or actions from alarms. Alerts are like a worker monitoring a group of alarms; when all the lights for the group are lit up, the alert is activated. Alerts can trigger action plans. See Control and scaling concepts
Action plan	A sequence of actions to perform on entities on the platform, such as VMs or scaling groups. An action plan is run by a trigger.
Trigger	A trigger is an alert or a schedule that will run the action plan, for example, during times of increased demand.
Scaling group	For horizontal autoscaling, create a scaling group for a VM with rules to define how the platform should scale it out. You can then include scaling operations in an action plan.
Vertical scaling	Vertical scaling means scaling up, adding more resources to an existing VM, for example, boosting your CPU and or RAM capacity.
Horizontal scaling	Horizontal scaling means scaling out, deploying more VMs when you need more resources.

Monitor virtual appliances

The monitoring system displays VM metrics and the user can configure alerts to trigger actions if certain conditions are met within virtual appliances. If your virtual datacenter supports monitoring, but you don't see any metrics for your virtual appliance, you may need to edit your VMs and configure monitoring on the VM monitoring tab. See VM monitoring and metrics

Privilege: Access virtual machine metrics

Then you can configure the display of metrics at the virtual appliance level.

To configure the refresh interval

Select the "Refresh data every" checkbox and enter a number of minutes.

Choose the metrics you wish to display and filter by metric statistics.

When you are fetching metrics, you can configure alarms on these metrics. See Manage cloud alarms

Manage scaling groups

To configure an automatic response to changing demands for resources, you can scale out VMs, which is also called horizontal autoscaling. To scale out, the platform clones the base VM and deploys the clones. To scale in, the platform will delete clone VMs and undeploy the base VM. Scaling operations are subject to all standard platform constraints, such as privileges and allocation limits.

Scaling notes

Limitations:
- The platform does not clone captured VMs, so you will need to create an instance and recreate the VM from the resulting template to use scaling groups. Create instances to save VM disks to templates
- VApp specs do not support scaling groups. See What do specs save and create
- You can push custom metrics for clone VMs but you cannot create alarms for cloned VMs that are part of a scaling group. This is because scaling groups have aggregate alarms that are associated with the base VM.
State of base VM: A scaling group with a deployed base VM would be destroyed if the base VM were deleted directly on the hypervisor. In contrast, a scaling group with an undeployed base VM is not vulnerable to interference at the hypervisor level

The platform enables you to automatically scale out (add more VMs) or scale up (add more resources to existing VMs).

Privilege: Manage scaling groups, Manage workflow for scaling groups

To use autoscaling do these steps:

Create a base VM, which can be deployed or undeployed
Configure the VM and enable metrics
Define a scaling group with rules for scaling the VM.
The checkbox to create a scaling action, will create the following automatically:
1. standard alarms and alerts for the selected metrics
2. action plans with scaling actions for the VM and triggers for the action plans, which are monitoring alerts
You can customize the elements the platform creates, or you can create your own configuration.

Define a scaling group

Before you begin:

Configure the base VM that will be scaled
Ensure that you have enough resources in your virtual datacenter to deploy up to the maximum number of cloned VMs, especially IP addresses

To create a scaling group:

Go to Virtual datacenters → Virtual appliances
Select the base VM, and from the menu on the VM icon, select the Create scaling group option
Enter the scaling parameters and rules
Click Save

Field	Description
Name	Name of the scaling group
Default cooldown	Period of time to wait from the start of one scaling operation before allowing another scaling operation
Minimum running virtual machines	The minimum running VMs that Abiquo must maintain in the scaling group. The value must be greater than or equal to zero, where zero means that the base machine is not deployed
Maximum running virtual machines	The maximum running VMs that Abiquo must maintain in the scaling group
Keep virtual machines in the same layer	Select this option to maintain VM anti-affinity layers when autoscaling
Disable workflow	Administrators with the privilege to Manage workflow for scaling groups can use this option to disable workflow or enable it as required.
Create in maintenance mode	Select this option to create the scaling group with maintenance mode ON, for example, to delay the automatic deployment of VMs to meet the minimum size
Create autoscaling action	Create basic operations to scale in and scale out, with triggers based on metrics and alarm conditions.
Scaling rules	Scaling rules describe how Abiquo will behave when a scaling event is triggered, for example, by an alert or a schedule.

Scaling rules

Field	Description
VMs	For a scale out rule, the number of times to clone the base machine and deploy each clone for each scaling step. For a scale in rule, the number of machines to remove from the scaling group (Abiquo will delete clone machines and undeploy the base machine)
Cooldown	The cooldown time to wait between scaling operations under this rule
Start	Date and time to start the time range when scaling triggers will start a scaling operation under this rule. The time range must be unique and cannot overlap with other rules with the same scaling direction. If there is no time range, then this is a default scaling rule.
End	Date and time that is the end of the time range when scaling triggers will start a scaling operation. The time range must be unique and cannot overlap with other rules with the same scaling direction. If there is no time range, then this is a default scaling rule

When you save the scaling group, Abiquo will mark the VM icon with the scaling group symbol and display the scaling group name. If the minimum number of deployed VMs are not present and the scaling group is not in maintenance, Abiquo will create clones of the base VM and deploy them to reach the minimum size. The number in the bottom right-hand corner of the icon is the number of running VMs in the scaling group, including the base VM.

To open the scaling group and check its parameters, click the scaling group symbol in the top right-hand corner of the VM icon.

Trigger autoscaling

Before you begin:

Create a VM and a scaling group for the VM. See Define a scaling group

To trigger autoscaling operations:

Create an action plan with a scaling action for the VM with the scaling group. See Create an action plan to automate VM actions
Create triggers to run the action plan. See Create a trigger for an action plan

When scaling, the platform will search for a scaling rule that is valid for the specific time range, or for a default rule. It will create or delete/undeploy the number of VMs in the rule, then wait for the cooldown period before accepting another scaling request.

How the platform scales out a VM

To scale out, the platform does not deploy VMs that are undeployed in the scaling group. To clone the base VM, the platform will do the following:

Create disks using the following:
1. Copies of content of disks from the VM template
2. Empty disks or volumes for each additional disk used in the VM
3. Disk controllers used in the VM
Apply ALL configuration used in the VM, for example:
1. CPU and RAM
2. Network connections of the same type (e.g. private network)
3. Assignment of firewall policies and attachment to load balancers
4. Chef recipes, backups, cloud-init, variables, and so on
5. Metrics. The group of metrics from clone VMs and the base VM (if it is deployed) can activate alarms in the base VM, even if it is not deployed
6. Exception – Alarms: the scaling group has only one set of alarms in the base VM

To scale in, Abiquo currently selects the VMs to delete or undeploy using first in, first out (FIFO). The platform deletes and undeploys VMs without requesting user confirmation when there are disks that are not stored in the Apps library (ISO configuration drive or additional hard disk).

Perform maintenance on a scaling group

To make changes to your VMs in a scaling group (manually deploy, undeploy, delete, etc.) and edit the scaling group, put it into maintenance mode, which will disable autoscaling. Note that you cannot delete the base VM without deleting the scaling group.

When you leave maintenance mode, the platform will apply your modifications to the scaling group, e.g. adding new rules. Then the platform will adjust the number of VMs in the group to within the minimum and maximum size range.

To manually put the scaling group in maintenance mode:

Go to Virtual datacenters → Virtual appliances → select VM
Click the "cog" maintenance button

To automatically put a scaling group in maintenance mode:

Trigger an action plan with the action "Scaling group: start maintenance mode"

To manually leave maintenance mode

Click the "cog" maintenance button again

To automatically leave maintenance mode

trigger an action plan with the action "Scaling group: end maintenance mode".

Note that although you can push custom metrics for clone VMs, you cannot create alarms for cloned VMs that are part of a scaling group. This is because scaling groups have aggregate alarms that are associated with the base VM.

Move a scaling group to another virtual appliance in the same virtual datacenter

To move a scaling group in maintenance mode to another virtual appliance in the same virtual datacenter:

Click the Move VM button
Select the new virtual appliance

Restrict a scaling group

To move a scaling group to a restricted virtual appliance, do these steps:

Go to Virtual datacenters → select scaling group
Click the VM move button on the VM control panel
Select the checkbox to Move to a restricted virtual appliance, select a restricted virtual appliance, or create a new VApp and select the Restrict virtual appliance checkbox

Delete a scaling group

When you delete a scaling group, the platform will place all the VMs in the virtual appliance as regular VMs and the scaling group constraints will no longer exist.

To delete a scaling group:

Go to Virtual datacenters → select scaling group
Click the "cog" button to put the scaling group into maintenance mode or run a maintenance action in an action plan
Click the delete button

Manage scaling groups with the API

API Documentation

For the Abiquo API documentation of this feature, see Abiquo API Resources and the page for this resource ScalingGroupsResource.

Pages related to monitoring and scaling groups

Main VM configuration page: Configure Virtual Machines
Configure VM Monitoring and Metrics