trade-tariff-lambdas-status-checks

Scheduled go lambda function to check the P90 status of the various services and store the results in an status.json file in s3 for review on a static status page

sequenceDiagram
    participant C as Cloudwatch Event
    participant L as Lambda Function
    participant S3 as S3 Bucket
    participant CF as Cloudfront Distribution
    participant U as User

    C->>+L: Triggers every 5 minutes
    L->>+L: Iterates through applications and check status
    L->>+S3: Drops status.json update file
    S3->>-CF: Serves index.html via Cloudfront
    CF->>U: User accesses status page
    U->>U: Sees status of each application

Loading

Individual application checks are configured per environment using a configuration type struct that is loaded from one of:

• config/applications.development.toml
• config/applications.staging.toml
• config/applications.production.toml

An example configured application for config/applications.development.toml

[[application]]
name = "Trade Tariff FPO Search"
url = "https://search.dev.trade-tariff.service.gov.uk/fpo-code-search"
authHeader = "X-Api-Key"
authHeaderValue = "env:API_KEY"
verb = "POST"
data = '{"description":"cheesecake"}'
errorHighWatermarkPercentage = 50.0
errorMediumWatermarkPercentage = 20.0
errorLowWatermarkPercentage = 1.0
p90HighWatermarkSeconds = 10.0
p90MediumWatermarkSeconds = 3.0
p90LowWatermarkSeconds = 1.0
concurrency = 3

The meaning of these configuration items is as follows:

Field Name	Description
Name	The name of the application configuration (seen in the status page).
url	The URL associated with the application, used for network requests.
authHeader (optional)	The key/name of the authorization header to use in network requests (e.g. X-Api-Key or Authorization).
authHeaderValue (optional)	The value to be used for the authorization header in network requests. (e.g. env:API_KEY). These are injected at deploy time
verb (defaults to "GET")	The HTTP method (verb) to be used in network requests, such as GET, POST, etc.
data (optional)	The data to be sent in the body of network requests, if applicable.
errorHighWatermarkPercentage	The high watermark percentage for error rates beyond which an application is considered to be in a INACTIVE state.
errorMediumWatermarkPercentage	The medium watermark percentage for error rates that indicate a DEGRADED but acceptable state.
errorLowWatermarkPercentage	The low watermark percentage for error rates considered ACTIVE under normal operation.
p90HighWatermarkSeconds	The high watermark for 90th percentile response times, above which performance is considered INACTIVE.
p90MediumWatermarkSeconds	The medium watermark for 90th percentile response times, indicating caution but not necessarily DEGRADED state.
p90LowWatermarkSeconds	The low watermark for 90th percentile response times, considered ACTIVE under normal conditions.
concurrency	How many workers to execute concurrently for this application.