Conversation
Introduces a CircuitBreaker class implementing the CLOSED → OPEN → HALF_OPEN state machine with configurable failure threshold and exponential backoff cooldown. Includes a CircuitBreakerRegistry for lazy per-DAO breaker creation and unit tests covering threshold, reset, and probe behavior. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Wires the CircuitBreakerRegistry into config (env vars), the proxy catch-all, the fan-out helper, and both service classes. Adds a global CircuitOpenError handler returning 503. Exposes per-DAO circuit states on the /health endpoint. Updates all affected tests. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Search endpoints were missing the cache-control middleware, causing search results to bypass the gateful Redis cache-aside layer entirely. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Sync dev with main
Adds routing score multipliers to nodeful upstreams across all three eRPC project configs to ensure nodeful is preferred over other upstreams. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…rity fix: prioritize nodeful upstream in eRPC dev config
…lity feat(gateful): add observability
Anticapture SDK npm
Feat/circuit breaker on gateful
…nticapture into fix/prometheus-indexer-alerts
This reverts commit f4f9b61.
Fix: prometheus indexer alert spam
Grafana "Anticapture V2" data visualization Dashboard
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 13e3d39ce9
ℹ️ About Codex in GitHub
Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".
| @@ -1,3 +1,4 @@ | |||
| #!/bin/sh | |||
| envsubst < /etc/prometheus/prometheus.yml.tmpl > /etc/prometheus/prometheus.yml | |||
| rm -rf /prometheus/* | |||
There was a problem hiding this comment.
Preserve Prometheus TSDB across container restarts
Removing /prometheus/* at startup wipes all persisted time-series data every time the Prometheus container restarts, which resets alert evaluation windows and destroys historical metrics needed for dashboards and incident analysis. In any deployment with a persistent volume mounted at /prometheus, this turns routine restarts or rollouts into guaranteed data loss; startup should not delete the TSDB directory.
Useful? React with 👍 / 👎.
No description provided.