fix: real CloudWatch alerting on both envs, add environment/release_version to logs (#160)
deploy/cloudwatch-alerts.sh had more bugs than the known log-group-name issue: wrong default region (eu-west-1), a metric-alarm call that would fail at runtime (--statistic and --extended-statistic are mutually exclusive, found by actually running it), a DB-connections alarm pointed at AWS/RDS that could never fire (this project uses self-hosted Postgres on EC2, not RDS - removed rather than left silently broken), and CloudWatch Agent install instructions using yum on Ubuntu instances.
Fixed all of it and ran the script for real against both staging and production: 30-day log retention set, SNS topics with both Gabriel and Aaron subscribed, 3 metric filters and 2 alarms (5xx rate, p99 latency) live on each environment. Verified via aws logs/cloudwatch describe-* directly, not just the script's own output.
Added environment/release_version to every log line (logging_config.py + a new Settings field + deploy.sh passing RELEASE_VERSION from the existing IMAGE_TAG) - verified end-to-end locally, takes effect on next deploy.
Semantic Changes
31 symbols
Files Changed
~6
1060 in snapshot
0 comments
muse hub commit comment sha256:7afc59b2ee9c70d18b748d269602c588385d6e936cc22b056fd775ab9f2fd499 --body "your comment"
No comments yet. Be the first to start the discussion.