Forgejo's web action routes don't support API token auth for private
repos (only session cookies or public access). Switch log fetching to
read the zstd-compressed log files directly from indri via SSH —
Forgejo stores all runner logs on disk regardless of which runner
executed the job.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
runner-logs now always authenticates with the Forgejo API token
(via --token flag, FORGEJO_TOKEN env, or 1Password) so it works on
private repos. The --repo default is auto-detected from the git
remote origin URL instead of being hardcoded.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The static asset cache block (css/js/png/etc) was missing
proxy_set_header Host, so Caddy received "forge.eblu.me" instead of
"forge.ops.eblu.me" and couldn't route the request. HTML loaded fine
because the main location / block had the header.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
All 7 ArgoCD containers had no resource limits, allowing them to consume
unlimited CPU/memory during node pressure events. This contributed to
cluster-wide probe timeout cascades on minikube-indri.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Comprehensive docs pass reflecting the new Fly proxy architecture:
- Fly proxy routes through Caddy on indri (not per-service TS Ingress)
- Direct WireGuard peering via --port=41641 pinning
- DERP relay performance lesson in Tailscale docs
- Caddy now in public traffic path
- indri tagged as flyio-target
- Removed fly-reload references
- Updated architecture diagrams and per-service setup guide
- Added changelog fragment
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Tailscale Ingress pods in k8s can't establish direct WireGuard
connections (stuck behind pod-network NAT → DERP relay → 20s latency).
Indri's host-level Tailscale CAN peer directly with Fly.
Change all nginx upstreams to route through Caddy on indri instead of
per-service Tailscale Ingress endpoints. Tag indri as flyio-target in
the Tailscale ACL so the Fly proxy can reach it.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Enable direct peer-to-peer WireGuard connections by pinning tailscaled
to port 41641 and exposing it as a UDP service. Without this, all
traffic routes through Tailscale DERP relays causing 20+ second
latency. Requires dedicated IPv4 (allocated: 168.220.82.221).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The nix-built Alloy image sets User=65534 (nobody). Even with
privileged: true, a non-root user gets no effective capabilities
(CapEff=0). Override with runAsUser: 0 so Beyla gets CAP_BPF and
CAP_SYS_ADMIN needed for eBPF instrumentation.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
NixOS defaults kernel.unprivileged_bpf_disabled=2, which blocks BPF
syscalls outside the init namespace even with CAP_BPF. Set to 1 so
privileged containers (Beyla/Alloy tracing) can create BPF maps.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Beyla (alloy-tracing) has been failing since April 13 with
"failed to set memlock rlimit: operation not permitted" because k3s
inherits the default 8MB memlock limit. Set LimitMEMLOCK=infinity on
the k3s systemd service so privileged containers can use eBPF.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The earthdistance extension (depends on cube) must be created before
restoring the teslamate database — discovered missing after 2026-04-13 DR.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Queries the Forgejo API to verify the target commit exists on the remote
before dispatching a build, preventing wasted CI runs on unpushed commits.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace Dockerfile with container.py for native Dagger builds.
Bump devpi-server 6.19.1→6.19.3, devpi-web 5.0.1→5.0.2.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
## Summary
- Replace per-request DNS resolution (variable-based `proxy_pass`) with static `upstream` blocks and `keepalive` connection pools
- Reuses TLS connections through the Tailscale tunnel instead of handshaking per request
- Add `mise run fly-reload` for nginx config reload without full redeploy (re-resolves upstream DNS)
## Trade-off
DNS is resolved at config load, not per-request. If Tailscale Ingress pods get new IPs (restart, reschedule), `mise run fly-reload` is needed. A Grafana alert will be added to detect this.
## Still TODO on this branch
- [ ] Grafana alert for upstream unreachable (triggers fly-reload reminder)
- [ ] Docs pass
- [ ] Deploy from branch and verify latency improvement
- [ ] Changelog fragment
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Reviewed-on: #337
Was sending Connection: upgrade on every proxied request, which is
semantically wrong for normal HTTP traffic. Use a map to conditionally
send 'upgrade' only when the client requests a WebSocket switch,
'close' otherwise.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Previous max bucket was 10s — all slower requests collapsed into +Inf,
making p50/p90/p99 unreadable during the Forgejo archive DoS.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Crawlers follow auth redirects to /user/login which is pointless for them.
Saves round-trips for both sides.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Points kustomization at v3.8.2-7a42aeb, the first image built from the
new container.py (replacing the Dockerfile).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The task was missing FLY_API_TOKEN injection, requiring manual fly auth
login. Now uses op read to fetch the deploy token from 1Password.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replaces the hand-written Dockerfile with container.py using the shared
alpine_runtime helper, which bumps the base image from Alpine 3.22 to 3.23.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The _1password-gui package doesn't export native messaging manifests
in the format the Firefox wrapper expects. The 1Password NixOS module
already handles native messaging host registration separately.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The xdg desktop entry and Librewolf user.js prefs didn't fix the
OAuth callback hang. Try stock Firefox instead as a simpler path.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The xdg desktop entry and mimeapps were already registered but
Librewolf doesn't delegate unknown URI schemes to the system
handler by default. This adds user.js prefs to complete the chain.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Facebook has been scraping forge mirror repos at ~3-4 req/s, slowing
down the Forgejo instance. Serve robots.txt directly from nginx to
disallow /mirrors/ while leaving eblume/* accessible to crawlers.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Shared Dagger helpers (src/blumeops/) affect all Dagger-built containers,
making path-based auto-triggers unreliable. All builds now go through
`mise run container-build-and-release <name>`.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Extend go_build() with buildmode and extra_env params, migrate miniflux
and forgejo-runner to use it, and bump all Alpine bases from 3.22 to 3.23.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Update kustomization image tags to the new container.py-built images
(v4.1.1-r1-2c483ce, v1.0.1-2c483ce).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace Dockerfiles with native container.py for both transmission and
transmission-exporter. Updates base images (Alpine 3.23, Python 3.14),
pins uv to 0.11.6 instead of :latest.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
LaunchAgents now call borgmatic directly at its mise-installed path
instead of routing through `mise x`, which triggered macOS TCC
permission dialogs (e.g. "mise wants to access Documents") that hung
headless sessions and caused backup failures.
Also adds `mise install` to the ansible role so borgmatic installation
is fully managed, and pins the version in both mise.toml and the role
defaults.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
## Summary
- Upgrade Prowler from 5.22.0 to 5.23.0
- Remove the `enumerate-images` init container workaround from `cronjob-image-scan.yaml`
- Use native `--registry` and `--image-filter` flags now that upstream fix (PR prowler-cloud/prowler#10470) is released
The init container was a workaround for prowler-cloud/prowler#10457 where `--registry` args weren't forwarded to the provider constructor. We wrote the fix, it was merged, and v5.23.0 includes it.
## Test plan
- [ ] Build new container (`mise run container-release prowler 5.23.0`)
- [ ] Update kustomization.yaml with new image tag
- [ ] Sync prowler ArgoCD app from branch
- [ ] Manually trigger image scan job and verify `--registry` works natively
- [ ] Verify CIS and IaC scan cronjobs still work
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Reviewed-on: #336
Removed Grafana from the control description — no Prowler finding
references it. Tightened scope to match actual usage (ArgoCD wildcard
RBAC mute). Added workflow-bot scoping note.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
## Summary
- Adds automated node-level verification to `review-compliance-reports`: kubelet file perms/ownership, kubelet config args, etcd CA separation, RBAC cluster-admin bindings
- Mutes the 14 MANUAL Prowler findings via new `manual-node-checks.yaml` mutelist file
- New `node-config-automated-verification` compensating control documents the approach
- Script fails loudly (red FAIL + verdict panel) if any check deviates from expected values
## Test plan
- [x] `mise run review-compliance-reports` — all 12 node checks PASS
- [x] Injected bad expected value (perms 400 vs actual 600) — FAIL rendered correctly
- [x] Fixed colon-in-binding-name bug (kubeadm:cluster-admins) with tab-separated jsonpath
- [ ] After merge: sync prowler mutelist ConfigMap and verify next scan shows 0 MANUAL findings
## Note
Prowler coverage is minikube-indri only — ringtail/k3s is a known gap tracked separately.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Reviewed-on: #335
## Summary
- Add native Dagger `container.py` for forgejo-runner (Go + Alpine runtime, static binary with CGO for SQLite)
- Update kustomization to point to local registry image (tag is placeholder until CI builds)
- Uses existing `clone_from_forge("forgejo-runner", ...)` mirror
## Test plan
- [x] `dagger call build --src=. --container-name=forgejo-runner` passes locally
- [ ] CI container build from branch succeeds
- [ ] Update kustomization tag to built image, deploy from branch via ArgoCD `--revision`
- [ ] Verify runner registers and picks up jobs
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Reviewed-on: #334
The lockfile bakes in devpi URLs — Dagger does a locked install, not
fresh resolution. This is the mechanism behind the cold-cache failure.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
