Stateful Node Details

The Stateful Nodes Overview gives you access to key information and summary data about a node. It offers insights into cost efficiency, as well as details about persisted network and storage resources.

Stateful Nodes Overview

The Stateful Node Overview tab gives you quick access to insights and summary data about the node. You can obtain an overview of your current cost efficiency status and an overview of your persisted network and storage resources.

1. In the Spot console, go to Elastigroup > Stateful Nodes.

2. In the list of stateful nodes, click a Node Name.

   

You can see the node status next to the name. It can be one of these statuses:

Node Summary

The Node Summary gives you insights into your node usage.

You can see statistics from the first of the month to the current date, as well as the last 7 or 30 days.

The widgets show the:

• VM Size: The VM size assigned to the node. Green is a spot VM, and blue is on-demand.
• Hours: Number of hours the VM has been used.
• Potential Costs: The cost calculated if the VM had been on-demand.
• Actual Costs: The cost of the VM using spot VMs.
• OD Costs: The cost of on-demand VMs used. This should be zero unless no spot VMs were available and a fallback to on-demand was required.
• Saved: The amount of money saved using spot VMs.
• Savings Percentage: The percentage of savings when using spot VMs compared to the potential cost if all on-demand VMs had been used.

You can also see a quick reference point for vital information about the stateful node.

• VM Name: The Azure VM name. Click the VM Name to load the Azure portal for this specific VM.
• Node ID: The identifier assigned to the node by Spot.
• VM Resource Group: The resource group of the VM.
• Description: The description you provided when the node was created.
• Region: The region defined on the node.
• VM Image: The image defined on the node. It is not shown when the node is defined with OS persistence.
• Availability Zones: The Azure availability zones that are defined on the node.
• Creation Date: The date the node was created.
• VNet Resource Group: Resource group defined on the virtual network of the node.
• Virtual Network: The virtual network of the primary network interface.
• Subnet: The subnet name defined for the primary network interface.
• Spot VM Sizes: The list of spot VM sizes defined on the node.
• On-demand VM Sizes: The list of on-demand VM sizes defined on the node.

Persisted Resources

You can see a summary of the node's persisted resources. The status icons show the node's type of persistence:

• Green: Persistency defined
• Gray: Persistency not defined

Network and Storage are not displayed when the node is in a Paused state.

• Network: You can see information about the network interfaces attached to the node.

  

• Storage: You can see information about the storage resources.

  

Node Activity

The node activity bar records the activity in the stateful node for the selected time range.

Click Details below a specific state:

• Paused: When the stateful node is paused, you can view the actions that occurred on that node until it was paused. You can see the attached or detached disks and the network interface.
• Recycled: In a recycled stateful node, you can view the history of actions, such as paused and resumed actions.
• Resume: Continues the stateful node using a new VM and the persisted resources.
• Import: You can view the steps of the import process of a stateful node.
• Stopped: You can view the stopped VM. It might take a few minutes for the stateful node to appear as stopped.

If an error occurs, the stateful node transitions into an error state. The information about the failed activity will appear in the activity’s details.

To view scheduled activities, click Scheduled Activities.

To create a new activity, click Schedule New Activity.

Monitoring

The Monitoring tab shows you information about the utilization of your stateful node. It gives you recommendations for modifying the preferred spot sizes list and helps you identify performance issues. You can view rightsizing recommendations and graphs with historical CPU and memory utilization data for the last hour, 6 hours, day, week, or 14 days.

Right Sizing Recommendations

Recommendations help you make better decisions about updating the VM list recommendation to update the list of preferred VM sizes. They are based on data for the last 14 days. The data is scanned once a day for each stateful node.

For downsizing recommendations, the highest CPU and memory utilization each hour are used. Recommendations are shown if the utilization is less than 40%. The size recommendations depend on the VM sizes that ran in the previous 12 hours (smallest vCPU number for CPU utilization and RAM size for RAM utilization). The recommendation is for the next smaller size level (unless it’s already the smallest, then it won’t recommend).

For upsizing recommendations, the average CPU and memory utilization per hour are used. Recommendations are shown if the utilization is above 80%. The size recommendations depend on the VM sizes that ran in the previous 12 hours (smallest vCPU number for CPU utilization and RAM size for RAM utilization). The recommendation is for the next larger size level (unless it’s already the largest, then it won’t recommend).

You can also see the recommendations in the Elastilog, which is updated once a day, and in the notification center.

Costs

The cost breakdown data provides a comprehensive overview of your stateful node’s costs. It is available through both the console and the API.

Daily Spend Breakdown

The compute lifecycle shows how much your compute cost was on each day of the selected time period. Each bar is broken down by spot, on-demand, and reserved instance cost for that day. The All Resources data shows the cost breakdown of compute, storage, and network resources.

Hover over a specific bar in the chart to see the:

• Date (format: mm/dd)
• Compute daily spend
• Network daily spend
• Storage daily spend
• Total spend for all resources

Summary

The donut chart summarizes overall costs within the selected time range.

Hover over the chart to see the costs by:

• Spot
• OD (On-Demand)
• Network
• Storage