- periodicMetrics function that queries queue stats and pings metrics
- New METRICS_PING_PERIOD variable for configuring ping frequency
- CloudWatch alarm to call periodicMetrics
- Docs
- Tests
Fixes#118
- delete pollQueue and switch processQueueItem to use SQS events
- add rate limiting based on sliding window using hits database
- test for rate limit pause
Issue #110
- 10 is probably too many times for a non-recoverable error - e.g. permanently
unreachable consuming client
- 3 is probably enough for an intermittent problem - e.g. temporary service
outage
- Ensure METRICS_URL gets passed to functions
- Ensure dev stack deployment has a good mock upstream service URL
- Increase bucket retention to 180 days
- Tweaks to client and README to make setting custom service URL easier
- Tweaks to metrics.md to reflect current functionality
- lib/metrics.js to wrap metrics pings
- Add pings to functions
- Mock endpoint that just logs POST requests, used for metrics in dev stack
- Tests
Fixes#47
- Move tests to live alongside functions
- Move mocks into a test-startup module, start testing processQueueItem
- Add npm start script
- Switch onchange to use polling, resolves some weirdness in windows
Issue #3
- Ensure AWS API Gateway can accept binary requests
- Add multipart form parsing, light input validation
- Expand options in client to include other API fields
- Increase S3 object expiry to 30 days
- Update tests
Fixes#1
- Add prettier, deploy:dev, and client NPM scripts
- bin/client.js as a start on a Hawk-enabled example service client
- Rework accept endpoint function
- Switch to POST requests
- Enforce Hawk authentication, with hardcoded dev credential option and
DynamoDB credentials
- Add Credentials table in DynamoDB
- Tests for Hawk in accept function
- README updates
Fixes#2
Les Orchard
Les Orchard
Barry Chen