Once an organization’s cloud usage reaches a good maturity level, all information and decision-making have to be distributed across all the owners.
Distributed Cost Governance is a process through which a cloud management platform provides recommendations on activities that can help optimize cloud resources, send alerts for spikes or unexpected costs, and distribute all kinds of information across an organization. In some cases, interventions are required, such as stopping a service that’s costing more than expected.
Empowering teams to have a certain level of autonomy brings in new problems for the finance teams: the information about resources and services is with the team working on the projects. Ideally, all teams should be engaged and made responsible for the costs they’re the cause for. Ownership is part of the AWS culture, and they have designed their products that way.
However, this process of distributing the data is challenging because it requires:
- Locating the correct owner.
- Providing them with up-to-date and accurate information.
- Using the appropriate channel to present the proposals.
- Collecting and recording the responses.
Let’s review each challenge:
- Finding the owner: In most cases, the recommendations or alerts are related to a given resource or application. With tags, it might be possible to identify who the owner of that particular item is. Cost Categories can help define filters such as Account ID, Service Types, and Tags to assign teams as owners.
- Staying up-to-date with the latest: As soon as the information is available, it should be sent out as cloud information gets stale fast. For alerts, this information should be presented in real-time to mitigate the impact of the detected issue.
- Providing accurate information: Based on trusted sources, such as the Billing report.
- Using the appropriate channel: For alerts, an online real-time means should be used, such as a Slack workspace. In other cases, the information could be presented summarized in emails or dashboards. The Cloud Management Platform should provide a mechanism to define the rules for using one channel or another.
- Collecting and Recording the responses: In the case of actionable items, like those automated recommendations requiring approval, the information should be sent back to the platform to execute the corresponding actions. For auditing purposes, the responses should also be registered.
Let’s review some of the most common use cases:
- Cost Variance: An agent monitors any significant deviation of the daily, weekly, or monthly costs based on the defined categories.
In case any of the defined thresholds are exceeded, an alert will be sent with the details to the defined Slack Channel or Microsoft Teams.
- Automated guardrails: When using Reserved Instances AutoPilot, the operation logs and the results should be stored and used as reports.
- Recommendation engine and automations: An agent monitors if some recommendations can be executed automatically, and notifies the corresponding channels with the status of each action.
For instance, it could be quite useful having notifications in a Slack channel with the default actions handy, such as Snooze alert or specific action based on the resource types and the recommendation.
As an organization’s cloud usage grows, more information about usage, events, or issues is generated.
A company’s capacity to provide in-time responses to all the events directly impacts cost control and risk mitigations. It’s essential to spread the information to distribute the job and empower all the teams to make decisions on their resources to increase their capacity to make decisions. That’s the culture of ownership.
Not all the detected issues should be sent out as alerts – it can cause alert fatigue. Having an efficient, distributed cost governance is the key when taking control of the situation by engaging the right team at the right moment. You get all that by using a Cloud Management Platform like Cloudwiry as the brain for Distributed Cost Governance.