AWS Dashboard

AWS Cost Summary

VMware Tanzu CloudHealth aggregates all cost information from accounts, services, and assets across your cloud infrastructure to provide a comprehensive cost summary in this report.

Navigate to Dashboards > Pulse > Cost Summary to view the Cost Summary dashboard.

Section 1: Cost Summary Customization Options

Filters

Use the following standard filter options to view the report as per your business needs.

• Accounts
• Billing Accounts
• AWS Services
• Charge Type
• Regions
• Organization specific Perspectives (Labeled with the letter P)

Projected Cost

Select one of the projected cost options from the dropdown.

Top List Count

Select a display limit for the number of projected costs.

Overall Budget

Select a overall budget from the dropdown. Categorized and amortized budgets and budgets for previous years are not available for selection.

Non-Recurring Cost toggle

Click the toggle button to view the non-recurring cost in the cost summary.

Section 2: Cost Summary by Perspective

Customize using Filters, Category, Projected Cost, or Overall Budget.

• Month-to-date costs compared to the same time last month.

• Projected cost for month compared to last month’s cost. Projection is based on a 30-day rolling analysis of your historical usage of services by service item, region, account, and Perspectives.

• If you have configured an overall budget (Setup > Governance > Budget), you can select it from the Overall Budget dropdown. The selected overall budget will appear in the Budget VS Projected cost for the current month.

Section 3: Top Projected Costs by Perspective Group

Customize using Filters, Category, Top List Count, or Display Non-Recurring Cost Toggle.

Cost trends by a specific Perspective or a combination of Perspective Groups, for example, costs for a specific owner within your production environment.

• Current month-to-date cost.
• Previous month cost at this time.
• Projected month-end cost.
• Nonrecurring costs, if configured.

Section 4: Top Projected Costs by Service

Customize using Filters, Category, Projected Cost, Top List Count, or Display Non-Recurring Cost Toggle.

Costs trends by service or service item. The By Service view shows cost trends for specific AWS or Azure services you use, for example, EC2, S3, or RDS. The By Service Item view expands the services into sub-costs to provide granular detail, for example, EC2 Compute, EC2 Compute, EC2 Transfer, or EBS Storage.

Section 5: Monthly Cost Allocation

Customize using Filters, Interval, Category, or Chart Type.

• Cost history for the last 13 months for a specific Perspective.
• Links to the Cost History report to help you analyze trends in greater detail.

To subscribe the AWS Cost Summary Pulse, click Subscribe > New Subscription at the top-right corner.

AWS Health Check Pulse

The AWS Tanzu CloudHealth Health Check report is an executive assessment of your cloud infrastructure that highlights cost savings, cloud governance points, and areas for potential security risks. Unlike other Pulses, it is intended as a high level report spanning all assets across all Perspectives. The total possible monthly savings reported includes all sections, except the Security Risk Exposure section.

Data Generation

The data for the report is updated once per day. The bottom of the report has the date of last generation. All sections have a drilldown link to the detailed real-time data, to easily access specific assets or reports. Because the report’s data is cached daily, report values will occasionally differ from the detailed data. Finally, the savings shown represents only assets from configured accounts.

Immediate Saving: RI Modification Recommendations

Using the RI Modifier Service, we loop through all possible time intervals to find the maximum possible savings for dealing with unutilized capacity. The intervals that we scan are: 1 week, month to date, 1 day, and 1 month.

Immediate Saving: Unused EBS Volumes

It’s easy to accumulate EBS Volumes that are no longer attached to any instance. This section reports on any EBS Volume whose state is ‘available’ and thus not used by any instance. It will report the number of volumes, the total size, and the list price of the volumes. The list price includes any PIOPS cost associated with the volumes.

Cloud Governance: Untagged Instances

We’ve made it really simple to identify EC2 instances that have not been tagged, and are more than likely not sanctioned by any business group. Common actions for these instances are to either fix the tags or terminate them. The total savings represents the total cost, month-to-date, of the matching instances.

Cloud Governance: Assets Not Allocated to Perspectives

Great Perspectives have few to no assets in the “Other” group. If assets are provisioned properly, and Perspectives are setup well (including a group for valid assets not appropriate for the Perspective), then this section is great at identifying Active EC2 Instances that were not provisioned properly, such as when a lone engineer spun up a random instance for some testing, but didn’t tag it properly and left it running. The savings for each Perspective represent the total cost, month-to-date, of the ungrouped EC2 Instances. The savings are not included in the overall savings of the report.

Operational Savings: Underutilized EC2 Instances

Getting the most out of your instances for operational efficiency is everyone’s goal, and this section makes it easy to find the under-performers. The default for Severely Underutilized is less than 35% utilization.  You can customize this section of the report by selecting an EC2 Instance Policy that you’ve defined with your Instance Rightsizing Policy.  The dollar amount to the right of the listed instances represents the Wasted Cost column for the current month’s Instance Rightsizing report.

Operational Savings: Underutilized EBS Volumes

Workloads can change over time, and the provisioned IOPS needed now might be different than last month. We find in-use volumes that have over-provisioned IOPS, and report on the number of volumes, the number of wasted PIOPS and the cost of those wasted PIOPS. Wasted PIOPS are calculated per volume as “PIOPS - (Max Volume Reads + Max Volume Writes).” If you have provisioned 4000 IOPS for a volume, and over the course of a week it only bursts to 2000 Reads/s and 1000 Writes/s, then 1000 are considered wasted.  You can customize this section of the report by selecting a Volume Policy that you’ve defined with your Volume Rightsizing Policy.

Security Risk Exposure

Instances are not double counted within this section. If an instance is included in “All Ports Open” it is not included in the count for “Unencrypted Ports.” The risk exposure reported represents the highest found in the sub-sections.

Security Risk Exposure: All Ports Open to the Public

A common best practice is to expose only ports that are needed to the outside world. This section helps uncover the most egregious abusers of security group rules: EC2 instances with all TCP/UDP/ICMP ports exposed to the public. Instances within a VPC are only included if they have a Public IP. Rules matching specific security groups are not included.

Security Risk Exposure: Unencrypted Public Port Exposure

We’ll admit it: the most common use case for this section is to yell at some engineer who opened up a port for some quick testing and forgot to close it.

We currently cover many common ports that handle unencrypted traffic by default, such as: FTP, Telnet, SMTP, POP3, NTP, IMAP4, SNMP, LDAP, SMTP, and RDP. Additionally, we scan some common databases: SQL Server Replication / Monitoring, Oracle, MySQL, Postgres, Cassandra Thrift Client, CouchDB, ElasticSearch, git, Redis, OrientDB, and MongoDB. If an instance matches multiple ports, it will only be counted once overall for the summary, but its count will be reported for each port. Instances within a VPC are only included if they have a Public IP. We also use the most popular name for a port; so a port like 11211 that is used by both memcached and the CouchBase Client Interface will have memcached listed. Please send us additional ports or applications to scan if you want them added to this list.

Security Recommendations

Security recommendations made by the AWS Security Best Practices default policy. The recommendations appear only if you have enabled the default policy. They help you understand what the particular security issue is and what action you can take to address it. For information on how to enable the default security policy, see The Need to Actively Manage Security.

Reservation Management Monthly Savings

The opportunity for the greatest monthly savings generally comes from diligent RI Management. We calculate potential savings using two budgets numbers. The first is using a limitless budget which gets us the fully optimized savings. We then use 20% of that as an initial budget for quick hits.

RI Utilization Pulse Summary

The Tanzu CloudHealth Reserved Instance Utilization Pulse is designed to provide you with a detailed operational and financial overview of your Reserved Instance (RI) environment. The goal is to enable you to better manage usage based on historical trends, identify opportunities for RI savings, and highlight changes you can make to optimize your compute costs. The Pulse is subscription based to facilitate sharing information with stakeholders and provide easy access to the data.

The RI Utilization Pulse is a compilation of several RI Management reports. The top of the report provides a summary of the health of your reserved instance usage for the current month to date (MTD). It provides you the total savings you have accumulated through your previous purchase of reserved instances, which are calculated based on running the equivalent instances with on-demand pricing.

Top N Compute Costs

The Top 10 Compute Cost by Group (MTD) provides cost and reservation savings by a selected business perspective (e.g. environment, application, department). You can configure the perspective for this report under Options. The Compute Cost column represents the aggregation of all instance charges for each group in the perspective. The Compute Cost if On Demand provides the compute costs you would have been charged if all usage in this group was on-demand. The Reservation Savings provides the total compute savings you have realized in this group over on-demand usage through the purchase of reserved instances.

Expiring Reservations

Expiring Reservations provides an overview of reservations that will be expiring in the next 90 days by type (e.g. All Upfront, Partial Upfront, No Upfront, etc). This allows you to have an at a glance view of upcoming expirations that could impact your future cost savings.

Top N Underutilized Reservations

Reservations get randomly applied across an account or series of linked accounts each hour, with an affinity for the account in which the reservation was purchased. This means that in order for a reservation to be used, a corresponding instance of the same instance type, availability zone and operating system, must be running and not already covered by a reservation. If an applicable running instance cannot be found for a reservation, it will go unused.

The Top N Underutilized Reservations (MTD) section of this report summarizes the specific reservations that have not been fully utilized month to date (MTD). Each row in this report will represent an underutilization of one or more reservations for a specific instance type, availability zone, operating system and reservation type. The Hours Unused provides the cumulative total of hours in which a reservation was not applied to a running instance. The Unrealized Savings summarizes the cost savings you could have realized if these reservations were fully utilized.

In particular you should pay close attention to underutilization of Heavy reservations, for which your are charged 100% usage irrespective of actual usage in a month.

Top N Reservation Opportunities

The Top N Reservation Opportunities (MTD) provides an overview of the top opportunities for you to save on compute costs through the purchase of reserved instances. This is sorted in order of the highest to lowest potential savings. Each row in the report provides a type of instance (instance type, availability zone, operating system) that is running on-demand. The Hours Used and Hours Reserved provides you details on the number of hours that instances of this type were used and reserved to date in the month. The Potential Savings allows you to quantify the maximum monthly savings you could achieve through the purchase of reserved instances to cover 100% of the on-demand usage (note: the maximum savings is based on the purchase of a 3-year All Upfront RI).

For consistently running workloads, you should optimize costs with the purchase of a reserved instance.

Top N Modification Opportunities

The Top 10 Modification Recommendations provides recommended changes to make to your existing reserved instance purchases to better utilize their cost savings. The recommendations are based on analyzing the month to date hourly usage of your reservations. Since modifications to existing RIs is a no-charge transaction with Amazon, it is highly recommended that you take advantage of the suggestions provided here. Each item in this section provides the exact change required to achieve the reported Monthly Savings. Typical changes involve moving a reservation between availability zones or changing its instance type within a family.

Instance Usage History

The Instance Usage History provides a visual chart that shows instance usage categorized by Reservation type. This will show usage of all instance types (e.g. Heavy, Medium), including on-demand and spot usage. This chart provides an excellent way to see trends within an environment and determine if there are opportunities for additional RI purchases to optimize cost. The chart defaults to showing daily data, but this can be configured under Options to show hourly, weekly and monthly as well.

Options

This report has several configuration options available, including the ability to change the number of items to show in a section (e.g. 10, 20), the time granularity of the Instance Usage History (e.g. hourly, monthly), and whether to limit the RI analysis to each account. The latter feature is useful for customers who purchase their reservations within each account and want to analyze their usage per account.