Skip to content

Consumption

The “Consumption” dashboard presents the organization’s daily consumption data on the provider, enabling a clear and detailed view of usage. The information is segmented by Workspace, Service, and SKU/usage category.

Consumption information is displayed by:

  • Usage date: date of resource consumption, regardless of when this usage was billed.

    • Loading and updates of billing data

    Data is collected on a daily basis at 08:00 (America/Sao Paulo).

    • Last update: Last time when billing data was loaded into our platform.

    Prerequisites

    1⃣ Service Account registered on the integrations section must have all necessary read permissions:

    On the BigQuery table containing the billing data export.

    On the bucket with the files containing the billing data.

    On the storage container that keeps the files with billing data.

    In the Tenancy usage reports (usage-reports).

    2⃣ Data must be loaded into the section financial structure.

    After integrating with our platform it is necessary to load your financial structure so the billing data can be properly processed.

    3⃣ Availability of the billing data for each cloud provider:

    Each provider has a specific timeframe for making billing data available; delays or failures directly impact the synchronization of data on the Platform's dashboards.

    • Billing data must be available on the BigQuery table that was chosen on the integration steps.

    Usually this data is sent to the BigQuery table with a delay of at most 24 hours after configuring the billing data export on GCP. Depending on the load and environment this can take more time. Further details can be found on the provider's documentation.

    • Files containing the billing data must be available on the appropriate storage buckets that where chosen on the integration process.

    After enabling the report it can take up to 24 hours until these files are exported to the S3 bucket. More information is avaible on the documentation of the provider.

    • Files containing the billing data export msut be available on the storage containers configured on the integration steps. Normally it takes 4 hours for the first data files to be available on the container storage from the moment the export is configure on the Azure Portal. More information about this process can be found on the provider's documentation.

    Cost reports are generally generated automatically every 6 hours. Although they contain hourly usage data, the data may have a delay of up to 24 hours to be fully consolidated. Synchronization on the platform depends on the generation of these files by Oracle in the billing bucket. Learn more in the provider documentation.

    Without integration

    When an user opts to skip the integration steps while acessing the platform for the first time an image is shown. This is to guide him through the necessary steps to load and visualize his data on the platform.

    Filters

    Filters are crucial to enable a detailed analysis of the data, highlighting specific information and easing informed decision making.

    • Usage period → Indicates the period in which the resource was consumed, therefore it is not possible to choose a future date.

    • Cost center → Cost centers are units that aggregate spending associated to specific activities or departments.

    • Cost center Manager → Name and e-mail of the manager responsible for the cost center.

    • Workspace→ Name and provider of the associated workspace.

    • GCP → Project Name
    • AWS →  AccountID
    • Azure → Subscription Name

    • OCI → Compartiment

    • Service → Name/Provider of the service.

    The period filter enables visualizing the costs of your enterprise in a specific interval, facilitating an improved comprehension of your resource consumption on the cloud.

    Beyond that, other filters offer options for a detailed analysis based on a cost center, manager, workspace or specific service. This promotes a segmented and strategic view of your financial data.

    It is possible to combine filter options to refine your analysis. Each selected filter adjusts the available options on the other filters automatically, providing the means to a focused search. For example: in selecting the cost center 'Financial' the filter for workspaces will show only the workspaces that are related to this cost center.

    Resetting filters

    You can clear all selected filter options using the 'Clear all' option, you also can clear specific options by clicking on the 'x' icon next to the selected filter.

    Consumption totals

    The total consumption is composed of all the credited and debited values throughout the period:

    Consumption

    Consumption

    Obs: The symbol “R$” in blue indicates credited values, while the “R$” in red represents debted values.

    • Resource consumption → Costs related to all resources being utilized.

    • Support → Costs related to the support fee of each cloud provider.

    • Commited Usage → Discounts related to continuous usage based on the resources.

    Important

    All negotiations related to commited usage must be made exclusively and directly with each cloud provider.

    • Credits → Credits estabilished through a contract with the cloud provider.

    • Total consumption → Value to be paid taking in consideration discounts, credits and adjustments.

    Time Zone (GCP)

    Cost data may show variations between the GCP Console and exported data due to differences in time zone handling:

    • GCP Console: Displays cost data using Pacific Time. During Pacific Standard Time (PST – winter), the time zone is UTC-8. During Pacific Daylight Time (PDT – summer), the time zone is UTC-7. The values shown in the console reflect this time zone.

    • Exported data: On the other hand, data exported from GCP follows the UTC (Coordinated Universal Time) standard. UTC is a time standard that serves as the basis for calculating time zones worldwide.

    When comparing GCP Console data with exported data, it is common to observe discrepancies, especially around day boundaries or in analyses with hourly granularity.

    Example 1 with Chart

    To illustrate the time zone difference and its impact on hourly and daily costs, see the chart below:

    grafico_fuso_horario_gcp

    Example 2

    1. Daily cost: Imagine that a resource generated a total cost of R$100.00 for an activity that occurred on 2024-01-20 at 02:00:00 UTC.
    • In the GCP Console (Pacific Time – PST, UTC-8 in winter): This cost will appear associated with 2024-01-19 at 18:00:00 (6 PM), since 02:00:00 UTC − 8 hours = 18:00:00 on the previous day. As a result, the total cost for 2024-01-20 and 2024-01-19 will differ depending on whether you view it in the GCP Console (Pacific Time) or in the cost table data (UTC).
    • In the exported data (UTC): The cost will be associated with 2024-01-20 at 02:00:00 UTC.

    As a result, the total cost of a given day in the GCP Console (Pacific Time) may differ from the total cost of the same day in the exported data (UTC).

    • Hourly difference: If you analyze costs by hour, the time zone difference becomes even more evident.
    • For example, a cost generated between 9 PM and 12 AM in São Paulo time (UTC-3) will be displayed in the GCP Console within the corresponding Pacific Time hours (UTC-8 or UTC-7).
    • However, in the exported data, this cost will be distributed across the corresponding UTC hours (12 AM–3 AM).

    Tips for interpreting the data correctly

    To avoid misinterpretation, follow these tips:

    • Know your time zone: Be clear about the Pacific Time (PST/PDT) used by the GCP Console and how it relates to the UTC used in the exported data.
    • Consistency: When comparing data, always use the same time zone reference, especially if you are using exported data (UTC).
    • GCP Documentation: Refer to the official Google Cloud Platform documentation for detailed information on how cost data is handled and which time zones are used in different contexts.

    Reports

    We offer report generation with a maximum period of 31 days, in CSV, making a detailed analysis of resources possible. For more informations visit the reports section.

    Charts - Consumption per workspace

    Exhibits the daily usage and distribuition of resource consumption per workspace.

    Workspace → Onicloud component that unifies the following concepts:

    • GCP

      Project → A GCP project is a container that organizes resources and related services, manages permissions and controls billing.

    • AWS

      Account → An AWS account is the level at which resources are provisioned, managed and organized, including multiple VPCs, projects and services.

    • Azure

      Subscription → An Azure subscription is a billing account where resources are provisioned, also where permissions and policies are managed.

    • OCI

      Compartment → The Compartment in OCI works as a logical container. For billing and cost integration purposes, the scope must be at the highest level.

    Consumption

    Shows daily consumption per service in dollars. Increasing the period range in the filter will make the date indicators adjust automaticaly, exhibiting a larger interval between them. The data points shown still correspond to daily usage idenpendent of the selected date interval. To visualize consumption details of a specific day you just need to place the mouse cursor over the it's corresponding point on the line in the graph. On all charts the top 10 workspaces with most cost are shown, the rest are unified under the "Others" tag.

    Consumption

    Shows the distribution of consumption per workspace, detailed by percentual in each one.

    Consumption

    Legends

    The graph legends appear in decreasing order, left to right, facilitating identification of the most expensive workspaces. Each appear listed by it's name and ID (except when it's an AWS workspace, where we show the ID two times because of the way the data is supplied by the provider)

    Clicking each name enables the exhibition of the complete name for that workspace. It also shows the "Others" category, which indicates the quantity of workspaces with costs in the given period but are not part of the top 10 most expensive ones.

    The other charts follow the same logic for exhibition.


    Service

    Shows daily consumption and usage distribuition per service via two graphs in timeline and in pizza format, like in the graphs for workspace.

    The original names are shown for each service in each provider, without changes. For example: the virtual machine services on GCP is called Compute Engine, on AWS it is called Elastic Compute Cloud.


    Usage category

    This term was unified to standardize the consumption view across different providers. While in GCP the concept is represented by the SKU (a unique identifier for product variation), in OCI, AWS, and Azure the term defines the specific product combined with its billing unit of measure (e.g., vCPU/hour or GB/month).

    This unification allows cost data to be displayed in a comparative and homogeneous format in our charts, just as with Workspace and Service.

    Anomaly Alerts

    Alerts notify you about unexpected and unusual spikes in your cloud service consumption, helping to avoid excessive and unplanned spending.

    How Anomaly Detection Works (The Logic)

    Our platform uses an intelligent statistics-based system to understand what is "normal" for your consumption. To do this, it defines a Maximum Acceptable Threshold for daily spending.

    If your spending from yesterday exceeds this threshold, and is also greater than a predefined minimum value, the alert is triggered.

    The Maximum Acceptable Threshold

    The system calculates the Maximum Acceptable Threshold by comparing your 30-day consumption with the variation from the last 15 days:

    • Historical Baseline (30 days): The system looks at your consumption over the last 30 days and finds the value that only the top 25% most expensive days exceed. We call this the High Historical Baseline.

    • Slack Range (15 days): The system measures how much your consumption varies over the last 15 days. This variation is multiplied by 1.5 (what we call "Slack") and is added to the High Historical Baseline.

    Alert Triggering

    The system marks an anomaly if:

    • Yesterday's Cost is greater than the Maximum Acceptable Threshold.
    • And the cost is greater than the Minimum Cost of R$ 300.00 (to avoid alerts about very small spikes).

    Example 01:

    Yesterday's cost > Maximum acceptable threshold

    and

    Yesterday's cost ≥ Minimum cost

    Anomaly

    Example 02:

    Metric Value Explanation
    High Historical Baseline (p75 30d) R$ 1,000.00 Only 25% of days in the last 30 days cost more than R$ 1,000.00.
    Recent Variation (15 days) R$ 100.00 The normal variation of your cost over the last two weeks is R$ 100.00.
    Slack (R$ 100.00 x 1.5) R$ 150.00 The slack is R$ 150.00.
    Yesterday's Cost R$ 1,250.00 The cost exceeded the Threshold of R$ 1,150.00. Alert Triggered.

    Accessing and Managing Alerts

    When an anomaly is detected, the event will be recorded on the platform and you will be notified by email.

    • On the Platform: You can access the Anomaly History directly in the consumption tab under SKU/Usage category.

    • Anomaly Management: The history screen allows you to:

      • View details such as the exact deviation value and which cloud service caused the spike.
      • Use filters to search for events by Provider (AWS, GCP, etc.) or Status.
      • Confirm Knowledge: When analyzing the anomaly, click Confirm Knowledge to change the status to Closed. This indicates that you are aware of the spending and prevents new alerts from being triggered about the same atypical consumption event.