Change:
- Improved error handling in buddy config
- Updated dependencies
Fix:
- Resolved issue with incorrect buddy timestamp
Fix:
- Buddy config unintended behavior, not ignoring own instance name from config file
- Buddy config unintended behavior, not using dashGoat 'updatekey' when none given for buddy.
- Some spelling
Change:
- Upgrade Go to v1.22.4
Change:
- Upgrade Go dependencies
- JSON return from API
Change:
- buddyDownStatusMsg to buddyDownStatus in config
Add:
- probeTimeoutStatus in config
Fix:
- Change time, after ttl timeout
Fix:
- Pagerduty timeout
Fix:
- Heartbeat change behavior
- Tests to check for change behavior
- Prometheus config error
Add:
- Pagerduty retry
Change:
- Low Pagerduty Timeout
- Upgrade Echo dependencies
Fix:
- Buddy state inconsistencies
- Buddy state race problems, making the app reset dataset
- Change timestamp not updated correctly
Change:
- Upgrade Go dependencies
Add:
- Config element logformat <txt/json>
- Heartbeat via HTTP GET (uri) /heartbeat/:urikey/:host/:nextupdatesec/:tags
- Alertmanager Webhook
- ack (acknowledge) to ServiceState
- Listen for Azure functions port
Fix:
- Missing version number on dashboard
- webpath prefix
Add:
- Slog logging
- PagerDuty push
- Report state change
Fix:
- Add missing buddynsconfig option to config file
Change:
- Upgrade Go + dependencies
- config element nsconfig is now called buddynsconfig
Deprecated:
- weblog config element
Fix:
- Logic error in relations to DisableMetrics
- Wrong state of timeoutprobe
Add:
- Basic Prometheus docs
Change:
- Change versioning to follow git tags
Fix:
- JS dashboard update error
- JS lowercase undefined error
Change:
- Made a common module, moved ServiceState to common for other tools
New:
- Embed webfiles to single binary
New:
- Prometheus /metrics endpoint.
- Metrics history with Prometheus backend, via /metricshistory/<host><service>/<hours>
- Metrics timeline for every service entry.
Change:
- New table layout with <div>
- Optimize DOM update.
Fix:
- Check for nextupdatesec, always beeing 19sec.
Change:
- ttlHousekeeping() to more readable code.
New:
- Add status favicon.
New:
- Add DependOn to reduce alert overload, depended services only show as info if source is down.
Change:
- Improved time translation to include days.
- Update 'Change' field behavior, adding timestamp when empty.
New:
- Native ENV config, instead of translation via Dockerfile.
- TtlOkDelete, seconds before deleting a service with state "ok".
Change:
- TTL behaviour, 4 config modes: Remove, PromoteOnce, PromoteOneStep, PromoteToOk (default).
New:
- k8s aware buddies and by extention, DNS aware buddies. DashGoat pods shoud be able to find each other, in the same namespace, provided there is a headless service called dashgoat-headless-svc.
Change:
- Code now under MPL 2.0
- Default port is now :2000
Change:
- Config code cleanup
Change:
- TTL feature, add option to only change state to ok, instead of beeing deleted
New:
- Buddy cluster feature, share state with your buddy instance over HTTP.
Will be added later.