The Banzai Cloud Cloudinfo application is a standalone project in the Pipeline ecosystem. While AWS, Google Cloud, Azure, AliBaba or Oracle all provide some kind of APIs to query instance type attributes and product pricing information, these APIs are often responding with partly inconsistent data, or the responses are very cumbersome to parse. The Cloudinfo service uses these cloud provider APIs to asynchronously fetch and parse instance type attributes and prices, while storing the results in an in memory cache and making it available as structured data through a REST API.
Building the project is as simple as running a go build command. The result is a statically linked executable binary.
make build
The following options can be configured when starting the exporter (with defaults):
build/cloudinfo --help
Usage of Banzai Cloud Cloudinfo Service:
--config-vault string enable config Vault
--config-vault-address string config Vault address
--config-vault-token string config Vault token
--config-vault-secret-path string config Vault secret path
--log-level string log level (default "info")
--log-format string log format (default "json")
--metrics-enabled internal metrics are exposed if enabled
--metrics-address string the address where internal metrics are exposed (default ":9090")
--listen-address string application listen address (default ":8000")
--scrape enable cloud info scraping (default true)
--scrape-interval duration duration (in go syntax) between renewing information (default 24h0m0s)
--provider-amazon enable amazon provider
--provider-google enable google provider
--provider-alibaba enable alibaba provider
--provider-oracle enable oracle provider
--provider-azure enable azure provider
--provider-digitalocean enable digitalocean provider
--config string Configuration file
--version Show version information
--dump-config Dump configuration to the console (and exit)
Create a permanent developer configuration:
cp config.toml.dist config.tomlRunning cloudinfo requires the web/ project to be built (requires Node.js to be installed):
cd web/
npm run build-prod
cd ..
build/cloudinfoThe cloudinfo service is querying the cloud provider APIs, so it needs credentials to access these.
Cloudinfo is using the AWS Price List API that allows a user to query product pricing in a fine-grained way. Authentication works through the standard AWS SDK for Go, so credentials can be configured via environment variables, shared credential files and via AWS instance profiles. To learn more about that read the Specifying Credentials section of the SDK docs.
The easiest way is through environment variables:
export AWS_ACCESS_KEY_ID=<access-key-id>
export AWS_SECRET_ACCESS_KEY=<secret-access-key>
cloudinfo --provider-amazon
Create AWS credentials with aws command-line tool:
aws iam create-user --user-name cloudinfo
aws iam put-user-policy --user-name cloudinfo --policy-name cloudinfo_policy --policy-document file://credentials/amazon_cloudinfo_role.json
aws iam create-access-key --user-name cloudinfo
On Google Cloud the project is using two different APIs to collect the full product information: the Cloud Billing API and the Compute Engine API.
Authentication to the Cloud Billing Catalog API is done through an API key that can be generated on the Google Cloud Console.
Once you have an API key, billing is enabled for the project and the Cloud Billing API is also enabled you can start using the API.
The Compute Engine API is doing authentication in the standard Google Cloud way with service accounts instead of API keys.
Once you have a service account, download the JSON credentials file from the Google Cloud Console, and set its account through an environment variable:
export GOOGLE_CREDENTIALS_FILE=<path-to-my-service-account-file>.json
export GOOGLE_PROJECT=<google-project-id>
cloudinfo --provider-google
Create service account key with gcloud command-line tool:
gcloud services enable container.googleapis.com compute.googleapis.com cloudbilling.googleapis.com cloudresourcemanager.googleapis.com
gcloud iam service-accounts create cloudinfoSA --display-name "Service account used for managing Cloudinfo”
gcloud iam roles create cloudinfo --project [PROJECT-ID] --title cloudinfo --description "cloudinfo roles" --permissions compute.machineTypes.list,compute.regions.list,compute.zones.list
gcloud projects add-iam-policy-binding [PROJECT-ID] --member='serviceAccount:cloudinfoSA@[PROJECT-ID].iam.gserviceaccount.com' --role='projects/[PROJECT-ID]/roles/cloudinfo'
gcloud iam service-accounts keys create cloudinfo.gcloud.json --iam-account=cloudinfoSA@[PROJECT-ID].iam.gserviceaccount.com
There are two different APIs used for Azure that provide machine type information and SKUs respectively. Pricing info can be queried through the Rate Card API. Machine types can be queried through the Compute API's list virtual machine sizes request. Authentication is done via standard Azure service principals.
Follow this link to learn how to generate one with the Azure SDK and set an environment variable that points to the service account file:
export AZURE_SUBSCRIPTION_ID=<subscription-id>
export AZURE_TENANT_ID=<tenant-id>
export AZURE_CLIENT_ID=<client-id>
export AZURE_CLIENT_SECRET=<client-secret>
cloudinfo --provider-azure
Create service principal with azure command-line tool:
cd credentials
az provider register --namespace Microsoft.Compute
az provider register --namespace Microsoft.Resources
az provider register --namespace Microsoft.ContainerService
az provider register --namespace Microsoft.Commerce
az role definition create --verbose --role-definition @azure_cloudinfo_role.json
az ad sp create-for-rbac --name "CloudinfoSP" --role "Cloudinfo" --sdk-auth true > azure_cloudinfo.auth
Authentication is done via CLI configuration file. Follow this link to learn how to create such a file and set an environment variable that points to that config file:
export ORACLE_TENANCY_OCID=<tenancy-ocid>
export ORACLE_USER_OCID=<user-ocid>
export ORACLE_REGION=<region>
export ORACLE_FINGERPRINT=<fingerprint>
export ORACLE_PRIVATE_KEY=<private-key>
export ORACLE_PRIVATE_KEY_PASSPHRASE=<private-key-passphrase>
# OR
export ORACLE_CONFIG_FILE_PATH=<config-file-path>
export ORACLE_PROFILE=<profile>
cloudinfo --provider-oracle
The easiest way to authenticate is through environment variables:
export ALIBABA_ACCESS_KEY_ID=<access-key-id>
export ALIBABA_ACCESS_KEY_SECRET=<access-key-secret>
export ALIBABA_REGION_ID=<region-id>
cloudinfo --provider-alibaba
Create Alibaba credentials with Alibaba Cloud CLI:
aliyun ram CreateUser --UserName CloudInfo --DisplayName CloudInfo
aliyun ram AttachPolicyToUser --UserName CloudInfo --PolicyName AliyunECSReadOnlyAccess --PolicyType System
aliyun ram AttachPolicyToUser --UserName CloudInfo --PolicyName AliyunBSSReadOnlyAccess --PolicyType System
aliyun ram CreateAccessKey --UserName CloudInfo
export DIGITALOCEAN_ACCESS_TOKEN=<access-token>
cloudinfo --provider-digitalocean
Create a new API access token on DigitalOcean Console.
Cloud providers can be configured one by one. To configure multiple providers simply list all of them and configure the credentials for all of them. Here's an example of how to configure three providers:
export AWS_SECRET_ACCESS_KEY=<secret-access-key>
export AWS_ACCESS_KEY_ID=<access-key-id>
export ALIBABA_ACCESS_KEY_ID=<access-key-id>
export ALIBABA_ACCESS_KEY_SECRET=<access-key-secret>
export ALIBABA_REGION_ID=<region-id>
export DIGITALOCEAN_ACCESS_TOKEN=<access-token>
cloudinfo --provider-amazon --provider-alibaba --provider-digitalocean
For a complete OpenAPI 3.0 documentation, check out this URL.
Here's a few cURL examples to get started:
curl -ksL -X GET "http://localhost:9090/api/v1/providers/azure/services/compute/regions/" | jq .
[
{
"id": "centralindia",
"name": "Central India"
},
{
"id": "koreacentral",
"name": "Korea Central"
},
{
"id": "southindia",
"name": "South India"
},
...
]
curl -ksL -X GET "http://localhost:9090/api/v1/providers/amazon/services/compute/regions/eu-west-1/products" | jq .
{
"products": [
{
"type": "i3.8xlarge",
"onDemandPrice": 2.752,
"cpusPerVm": 32,
"memPerVm": 244,
"gpusPerVm": 0,
"ntwPerf": "10 Gigabit",
"ntwPerfCategory": "high",
"spotPrice": [
{
"zone": "eu-west-1c",
"price": 1.6018
},
{
"zone": "eu-west-1b",
"price": 0.9563
},
{
"zone": "eu-west-1a",
"price": 2.752
}
]
},
...
]
}
1. The API responses with status code 500 after starting the cloudinfo app and making a cURL request
After the cloudinfo app is started, it takes a few minutes to cache all the product information from the providers.
Before the results are cached, responses may be unreliable. We're planning to solve it in the future. After a few minutes it should work fine.
2. Why is it needed to parse the product info asynchronously and periodically instead of relying on static data?
Cloud providers are releasing new instance types and regions quite frequently and also changing on-demand pricing from time to time.
So it is necessary to keep this info up-to-date without needing to modify it manually every time something changes on the provider's side.
After the initial query, the cloudinfo app will parse this info from the Cloud providers once per day.
The frequency of this querying and caching is configurable with the --product-info-renewal-interval switch and is set to 24h by default.
3. What happens if the cloudinfo app cannot cache the AWS product info?
If caching fails, the cloudinfo app will try to reach the AWS Pricing List API on the fly when a request is sent (and it will also cache the resulting information).
4. What kind of AWS permissions do I need to use the project?
The cloudinfo app is querying the AWS Pricing API to keep up-to-date info
about instance types, regions and on-demand pricing.
You'll need IAM access as described here in example 11 of the AWS IAM docs.
If you don't use Prometheus to track spot instance pricing, you'll need to be able to access the spot price history from the AWS API as well with your IAM user.
It means giving permission to ec2:DescribeSpotPriceHistory.
5. What is the advantage of using Prometheus to determine spot prices?
Prometheus is becoming the de-facto monitoring solution in the cloud native world, and it includes a time series database as well.
When using the Banzai Cloud spot price exporter, spot price history will be collected as time series data and
can be queried for averages, maximums and predictions.
It gives a richer picture than relying on the current spot price that can be a spike, or on a downward or upward trend.
You can fine tune your query (with the -prometheus-query switch) if you want to change the way spot instance prices are scored.
By default the spot price averages of the last week are queried and instance types are sorted based on this score.
6. What happens if my Prometheus server cannot be reached or if it doesn't have the necessary spot price metrics?
If the cloudinfo app fails to reach the Prometheus query API, or it couldn't find proper metrics, it will fall back to querying the current spot prices from the AWS API.
Copyright (c) 2017-2019 Banzai Cloud, Inc.
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.