[Alpha] AWS Lambda feature server

Warning: This is an experimental feature. It's intended for early testing and feedback, and could change without warnings in future releases.

To enable this feature, runfeast alpha enable aws_lambda_feature_server

Overview

The AWS Lambda feature server is an HTTP endpoint that serves features with JSON I/O, deployed as a Docker image through AWS Lambda and AWS API Gateway. This enables users to get features from Feast using any programming language that can make HTTP requests. A local feature server is also available. A remote feature server on GCP Cloud Run is currently being developed.

Deployment

The AWS Lambda feature server is only available to projects using the AwsProvider with registries on S3. It is disabled by default. To enable it, feature_store.yaml must be modified; specifically, the enable flag must be on and an execution_role_name must be specified. For example, after running feast init -t aws, changing the registry to be on S3, and enabling the feature server, the contents of feature_store.yaml should look similar to the following:

project: dev
registry: s3://feast/registries/dev
provider: aws
online_store:
  region: us-west-2
offline_store:
  cluster_id: feast
  region: us-west-2
  user: admin
  database: feast
  s3_staging_location: s3://feast/redshift/tests/staging_location
  iam_role: arn:aws:iam::{aws_account}:role/redshift_s3_access_role
flags:
  alpha_features: true
  aws_lambda_feature_server: true
feature_server:
  enabled: True
  execution_role_name: arn:aws:iam::{aws_account}:role/lambda_execution_role

If enabled, the feature server will be deployed during feast apply. After it is deployed, the feast endpoint CLI command will indicate the server's endpoint.

Example

After feature_store.yaml has been modified as described in the previous section, it can be deployed as follows:

$ feast apply
10/07/2021 03:57:26 PM INFO:Pulling remote image feastdev/feature-server:aws:
10/07/2021 03:57:28 PM INFO:Creating remote ECR repository feast-python-server-key_shark-0_13_1_dev23_gb3c08320:
10/07/2021 03:57:29 PM INFO:Pushing local image to remote 402087665549.dkr.ecr.us-west-2.amazonaws.com/feast-python-server-key_shark-0_13_1_dev23_gb3c08320:0_13_1_dev23_gb3c08320:
10/07/2021 03:58:44 PM INFO:Deploying feature server...
10/07/2021 03:58:45 PM INFO:  Creating AWS Lambda...
10/07/2021 03:58:46 PM INFO:  Creating AWS API Gateway...
Registered entity driver_id
Registered feature view driver_hourly_stats
Deploying infrastructure for driver_hourly_stats

$ feast endpoint
10/07/2021 03:59:01 PM INFO:Feature server endpoint: https://hkosgmz4m2.execute-api.us-west-2.amazonaws.com

$ feast materialize-incremental $(date +%Y-%m-%d)
Materializing 1 feature views to 2021-10-06 17:00:00-07:00 into the dynamodb online store.

driver_hourly_stats from 2020-10-08 23:01:34-07:00 to 2021-10-06 17:00:00-07:00:
100%|█████████████████████████████████████████████████████████████████| 5/5 [00:00<00:00, 16.89it/s]

After the feature server starts, we can execute cURL commands against it:

$ curl -X POST \                                 
    "https://hkosgmz4m2.execute-api.us-west-2.amazonaws.com/get-online-features" \
    -H "Content-type: application/json" \
    -H "Accept: application/json" \
    -d '{
        "features": [
            "driver_hourly_stats:conv_rate",
            "driver_hourly_stats:acc_rate",
            "driver_hourly_stats:avg_daily_trips"
        ],
        "entities": {
            "driver_id": [1001, 1002, 1003]
        },
        "full_feature_names": true
    }' | jq
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  1346  100  1055  100   291   3436    947 --:--:-- --:--:-- --:--:--  4370
{
  "field_values": [
    {
      "fields": {
        "driver_id": 1001,
        "driver_hourly_stats__conv_rate": 0.025330161675810814,
        "driver_hourly_stats__avg_daily_trips": 785,
        "driver_hourly_stats__acc_rate": 0.835975170135498
      },
      "statuses": {
        "driver_hourly_stats__avg_daily_trips": "PRESENT",
        "driver_id": "PRESENT",
        "driver_hourly_stats__conv_rate": "PRESENT",
        "driver_hourly_stats__acc_rate": "PRESENT"
      }
    },
    {
      "fields": {
        "driver_hourly_stats__conv_rate": 0.7595187425613403,
        "driver_hourly_stats__acc_rate": 0.1740121990442276,
        "driver_id": 1002,
        "driver_hourly_stats__avg_daily_trips": 875
      },
      "statuses": {
        "driver_hourly_stats__acc_rate": "PRESENT",
        "driver_id": "PRESENT",
        "driver_hourly_stats__avg_daily_trips": "PRESENT",
        "driver_hourly_stats__conv_rate": "PRESENT"
      }
    },
    {
      "fields": {
        "driver_hourly_stats__acc_rate": 0.7785481214523315,
        "driver_hourly_stats__conv_rate": 0.33832859992980957,
        "driver_hourly_stats__avg_daily_trips": 846,
        "driver_id": 1003
      },
      "statuses": {
        "driver_id": "PRESENT",
        "driver_hourly_stats__conv_rate": "PRESENT",
        "driver_hourly_stats__acc_rate": "PRESENT",
        "driver_hourly_stats__avg_daily_trips": "PRESENT"
      }
    }
  ]
}