The Metacloud Dashboard displays metrics about about activity and resource usage for virtual and hardware components in your cloud. If you are an administrator, you can view metrics for all projects within your deployment. If you are a non-administrative user, you can view metrics for your own project.
Viewing Metrics for a Project
After logging into the Dashboard, click Project in the left navigation bar. Then, click Overview.
The piecharts in the Limit Summary area provide a quick glance at overall project usage for certain resources.
- If the number of instances or virtual CPUs (VCPUS) used approaches full capacity, you may want to consider adding more to scale with the demands of the project's end-users.
- If the amount of RAM consumed approaches full capacity, you may want to consider allocating more, either by using higher-RAM flavors for your instances, or increasing the RAM for flavors currently assigned to the instances. See Managing Flavors.
- If you notice a sustained increased in network latency, a high number number of security groups in use may be a factor. Verification of compliance with multiple, complex rules in security groups can slow down traffic. In this case, you may want to consider reducing the number or complexity of the security groups in use. See Configuring Access and Security for Instances.
TIP: Another way to manage security groups for better performance is to enforce security group quotas. See Viewing and Managing Quotas.
The Usage table shows the breakdown of resource allocation by instance. You can click the name of an instance to see information about it, including whether it is active. If a given instance is inactive, and you determine it is no longer necessary, you can delete it and reallocate its RAM or CPU.
The middle area of the Overview tab shows total hours of usage for VCPUS, disk space and RAM. This is useful if you bill for cloud services. By changing the date ranges, you can isolate periods of increased usage hours, which can be helpful for capacity planning.
TIP: You can distribute usage metrics to stakeholders in your organization who don't have access to your Dashboard. Click Download CSV Summary to capture the data in a spreadsheet-friendly format.
Viewing Metrics Across All Projects in Your Cloud
If you are a Metacloud administrator, you can view metrics for all projects in your cloud. After logging into the Dashboard, click Admin in the left navigation bar. Then, click Overview.
As with the project overview, you can select a time period to view overall usage hours for VCPUS, disk space, and RAM.
The Usage table shows a breakdown of these hours by project and overall resource allocation by project. Comparing allocation for each tenant can be helpful for decision-making about how to redistribute resources as the needs of tenants change.
Viewing Metrics for Controllers and Hypervisors
If you are a Metacloud administrator, you can view metrics for controllers and hypervisors. After logging into the Dashboard, click Controllers or Hypervisors in the left navigation bar.
The Controllers tab displays usage graphs for each Metacloud Control Plane (MCP). All API endpoints reside on controllers, so if you are encountering API issues, you may find it helpful to check this tab. By changing the time setting in the top right corner of the display for any controller, you can isolate activity in windows ranging from the past hour to the past year. Viewing longer periods is useful for tracking usage trends toward capacity planning. If you see usage spike or increase over a period of time, you may want to consider expanding capacity.
Click VIEW LIVE STATS to see current, granular information about a controller's system, running processes, network traffic activity, and disk usage. This information updates every five seconds.
The Hypervisors tab displays pie charts for VCPUs, memory, and disk usage for each Metacloud Hypervisor (MHV). This is a quick way to learn if these levels are approaching capacity.
As for controllers, the Hypervisors tab displays usage graphs with adjustable time settings for each Metacloud Control Plane, as well as Live Stats. Hypervisors host instances, so if a given instance is unresponsive or slow, you may find useful, relevant data here about the cause of the issue. For example, a sustained spike in network usage may help explain latency.