How can I retrieve usage and cost data for my IBM Cloud account using a REST API? I found that there are billing related commands and I can export some data as JSON. Is there an API or SDK I can use, ideally Python?
Here are some of the IBM Cloud billing commands I use:
ibmcloud billing resource-instances-usage --json
ibmcloud billing account-usage --json
Are there equivalent APIs for that?
UPDATED: An API is now documented here: https://cloud.ibm.com/apidocs/metering-reporting
OLD:
I couldn't find a documented API but used the tracing to see how the above commands are executed. Using a valid access_token a program can call the metering host and obtain usage data for an account, resource group or all resource instances:
A GET on the following URL with an account ID and month as YYYY-MM returns a JSON object with all resource usage and related cost:
https://metering-reporting.ng.bluemix.net/v4/accounts/account_id/resource_instances/usage/?_limit=100&_names=true
I coded up a small Python script that dumps that data or prints it as CSV.
def processResourceInstanceUsage(account_id, billMonth):
METERING_HOST="https://metering-reporting.ng.bluemix.net"
USAGE_URL="/v4/accounts/"+account_id+"/resource_instances/usage/"+billMonth+"?_limit=100&_names=true"
url=METERING_HOST+USAGE_URL
headers = {
"Authorization": "{}".format(iam_token),
"Accept": "application/json",
"Content-Type": "application/json"
}
response=requests.get(url, headers=headers)
print ("\n\nResource instance usage for first 100 items")
return response.json()
The GitHub repo openwhisk-cloud-usage-samples uses a serverless approach to getting data via APIs. Examples are included in the repo. It's written in Javascript, but a package it uses openwhisk-jsonetl was designed so that you could declare the URLs and parameters in YAML (rather than writing code) to request and transform JSON.