Two years ago we published our first blog titled Infracost - cloud costs for devs. Infracost is now used by thousands of engineers across many CI/CD systems, 3rd party integrations and even VS Code, so they can see how much their code changes will cost before resources are launched.
Today I'm excited to announce Infracost Cloud - our SaaS product that builds on top of Infracost open source and helps team leads, managers and FinOps practitioners see all code changes and their cost impact from a central place. This creates an immediate feedback loop of upcoming changes and costs across the organization.
Currently companies receive billing alerts after-the-fact. But this is too late as the money has already been spent and the engineers have moved-on to work on other tasks (and rework is painful). Infracost Cloud gives team leads a tight feedback loop: you can review cost estimates from a central place before they hit the bill.
The dashboard below shows all pull requests over the last 30 days. Each dot is a pull request, dots above the 0-line will increase your costs, and dots below the line will decrease your costs. Dots with the same color are from the same project. You can quickly spot outliers and click on those dots to jump into pull request comments to help guide the team. This saves you money as you can catch issues before they hit your bill. It also gives you a way to provide positive affirmations to the team and acknowledge their cost saving efforts.
Currently it's time consuming for FinOps practitioners to find the cause of a spike or drop in cloud costs as you have to slice-and-dice the bills based on tags, then ask engineering teams if they changed something during the target date range. Infracost Cloud lets FinOps practitioners zoom-in on the target date range and see for themselves if the change was caused by infrastructure changes or user traffic.
This saves you time as you can correlate pull requests that increase or decrease costs the most, by how much, and by whom, so you quickly find the right people to talk to. You can also see the context around the pull request, for example the title and commit message, so you can ask better questions when talking to engineering teams.
Custom price books
By default, the Infracost CLI uses the public prices from AWS, Azure and Google. Infracost Cloud can be used to setup custom price books by applying discounts as part of AWS Enterprise Discount Program (EDP), Azure Enterprise Agreements, and Google Commitment Agreement. Once set, the discounts are used in all CLI outputs, pull requests and Infracost Cloud estimates.
We're continuing our investment in the open source Infracost Community edition and have a lot more planned. Upgrade to the recently released v0.10.9 version to use the following new features.
VS Code extension
The Infracost VS Code extension shows cost estimates for Terraform right in your editor. This has been updated to include a tree view so you can see which projects, files and blocks have the most impact to your costs.
Private registry modules
The Infracost CLI can now automatically download private registry modules hosted on Terraform Cloud or GitLab so you can see cost estimates for module blocks using those as their
Instances running less than a month
The Infracost usage files now supports a
monthly_hrs option for AWS, Azure and GCP virtual machines so you can get cost estimates for instances that run less than a month.
JSON Terraform var files
--terraform-var-file flag now also accepts JSON files when running
infracost breakdown or
diff so you can specify variable files similar to Terraform's
Unlike GitHub, GitLab and Azure Repos, pull request comments in Bitbucket do not support collapsible sections, which can make cost estimate comments quite long. To workaround this issue, we added the following options that exclude the CLI output so you can change the Bitbucket comment to only have the short summary table:
- a new
--exclude-cli-outputflag has been added to
- a new
bitbucket-comment-summaryoutput format has been added to
The following resources were added: Azure IoT Hub (
azurerm_iothub_dps) and Virtual Network peering (
azurerm_virtual_network_peering), AWS RDS
oracle-ee-cdb engine types.
Many thanks to the following first time contributors: @philharle, and @jessecureton, @sawyerw-mastery, @jauderho. Also shout-out to the following people for sharing their Infracost knowledge with the rest of the community - InfraSocks are coming your way!
- Seif Bassem: Estimate your Infrastructure-as-code costs before deploying
- Pete Mallam: Azure Pipelines, InfraCost and the dreaded 403 error
- Rahul Soni: Cost Monitoring & Insights with Kubecost & Infracost
- Pedro H. Santos: Descobrindo os custos da sua infraestrutura utilizando Infracost
- Marcon Willian: Prevendo Custo de Arquitetura Terraform
Finally, our all-remote team was in Edinburgh, Scotland for our first In-Person-Week. This is different to our Recurring Employee Meetup (REM) as we just show-up to an office for a week and do our regular work. AWS put the cherry on our in-person-week cake by featuring Infracost in their inaugural Build On AWS Twitch livestream!