1# Security Hardening Controls
2
3This page summarizes Gordon's security-related configuration knobs and the defaults operators should know before exposing an instance.
4
5## Registry upload quotas
6
7Registry blob uploads are bounded by two limits under `[server]`:
8
9```toml
10[server]
11max_blob_chunk_size = "95MB"
12max_blob_size = "1GB"
13```
14
15- `max_blob_chunk_size` limits a single upload chunk.
16- `max_blob_size` limits the cumulative blob/layer upload size.
17- Exceeding either limit returns an OCI-compatible size error and the failed upload is cleaned up.
18
19## Admin logs permission
20
21Container and deploy logs can include environment-derived output. Gordon gates log access behind a dedicated scope:
22
23```bash
24gordon auth token generate --subject ops --scopes "admin:logs:read" --expiry 30d
25```
26
27- `/admin/logs`, app failure diagnostics, and deploy failure logs require `admin:logs:read`.
28- `admin:status:read` does not grant log access.
29- Operation errors returned to `admin:apps:read` callers are stable and log-free. Application diagnostics are a separate field, redacted of the affected service's resolved secret values before storage, and returned only to callers holding `admin:logs:read`. If a declared secret cannot be resolved for redaction, Gordon drops the diagnostics instead of storing unredacted output. Mutation responses never include application output.
30- Process and container log sources remain operator-owned data. Gordon redacts common credential patterns when serving logs through its API, but this is defense in depth rather than proof that arbitrary application output is secret-free. Restrict filesystem, journal, runtime, backup, and `admin:logs:read` access accordingly.
31
32## Volume pruning scope
33
34Volume pruning removes a volume only when Gordon has a durable `released` ownership record, the runtime app/incarnation/service labels agree with that record, and no container mounts it. A `gordon.managed=true` label by itself is not deletion authority; retained, unknown, contradictory, and unrelated volumes survive.
35
36Use `gordon volumes prune --dry-run` before deletion. Do not substitute `docker volume prune` or an equivalent runtime command: runtime-native pruning bypasses Gordon's ownership and retention checks.
37
38Use dedicated admin scopes:
39
40- `admin:volumes:read` for listing volumes.
41- `admin:volumes:write` for prune operations.
42
43## Administrative app bind mounts
44
45App manifests cannot name host paths. A host bind exists only when the operator declares a named policy in `gordon.toml`:
46
47```toml
48[app_mounts.app-logs]
49source = "/srv/gordon/host-logs"
50read_only = true
51allowed_apps = ["metrics-agent"]    # exact, non-empty
52allowed_services = ["web"]          # exact, non-empty
53root = "/srv/gordon"                # optional boundary
54```
55
56The manifest references only the policy name: `[[service.bind]] name = "app-logs"` with an absolute container `path`.
57
58- `allowed_apps` and `allowed_services` are exact, non-empty allowlists, so least privilege is enforced by construction. There is no wildcard, prefix, or empty-means-all form.
59- `source` is resolved through symlinks and must be a regular file or directory under `root`. When `root` is omitted, the source's parent directory is the boundary. Devices, sockets, FIFOs, and escaping symlinks are refused.
60- Read-only precedence: `read_only` on the policy or `readonly` on the manifest bind forces the mount read-only; a manifest never weakens its policy.
61- Destinations must be absolute, normalized, and outside reserved container paths (`/`, `/proc`, `/sys`, `/dev`, `/boot`, and their children).
62- The policy is re-resolved immediately before every container create, restart, and recovery. Removing a policy or an allowlist entry blocks future deploys; running containers keep their current mounts until redeployed.
63- `gordon serve` reload republishes validated policies atomically. An edit that fails validation is rejected and the previous policies stay live.
64- Gordon never creates, deletes, chowns, backs up, or prunes the host path; ownership, permissions, and backup remain the operator's responsibility.
65
66## Administrative app devices (CDI)
67
68App manifests cannot name host devices. A device grant exists only when the operator declares a named policy in `gordon.toml`:
69
70```toml
71[app_devices.transcode-gpu]
72cdi = ["example.com/gpu=GPU-device-uuid"]
73allowed_apps = ["video"]          # exact, non-empty
74allowed_services = ["transcoder"]  # exact, non-empty
75```
76
77The manifest references only the logical name: `devices = ["transcode-gpu"]`.
78
79- `cdi` holds explicit CDI device IDs. Raw `/dev` paths, unqualified names, and aggregate `=all` selectors are rejected.
80- `allowed_apps` and `allowed_services` are exact, non-empty allowlists, so least privilege is enforced by construction. There is no wildcard, prefix, or empty-means-all form.
81- Gordon resolves logical names to CDI IDs at activation time and encodes them as one native CDI `DeviceRequest`. Revisions persist the logical names, never the host resolution.
82- The grant is re-resolved immediately before every preflight, container create, restart, and recovery. Revoking a grant blocks future deploys while the running service is untouched; changing a mapping never recreates a running container.
83- Device-bearing creates require Podman 5.4+ or Docker 28.3+ with native CDI configured. Older or unrecognized engines fail with a structured `runtime-unsupported` error and never run without devices.
84- Gordon installs no drivers, manages no quotas, and injects no device environment: images carry their own runtime expectations.
85
86## Pass import plaintext handling
87
88With the `pass` backend, Gordon imports eligible plaintext `.env` files at startup. It removes a source file only after every entry is stored successfully. If a destination entry already exists or an import fails, Gordon fails closed and leaves the plaintext source in place for operator review.
89
90## External image registries
91
92Docker Hub, `ghcr.io`, `quay.io`, and Gordon's configured registry are always allowed. Add every other registry hostname and non-default port explicitly:
93
94```toml
95[images]
96allowed_registries = ["registry.internal:5000"]
97require_digest = true
98```
99
100- `docker.io` and Docker Hub's canonical pull host `registry-1.docker.io` are equivalent.
101- Other registries are accepted only when their exact hostname+port is configured. Allowlisting does not configure credentials; external resolution and pulls are anonymous unless the runtime already has suitable access.
102- `require_digest = true` requires every image, including Gordon registry images, to use `@sha256:<64 hex chars>`.
103- The allowlist restricts hostnames, not resolved IPs. It cannot prove a DNS hostname is non-private and does not replace firewall or runtime egress controls.
104
105## Smart TCP Raw Fallback
106
107Raw TCP fallback on a `smart_tcp` edge entrypoint is disabled by default. Unknown non-HTTP/non-TLS bytes reach a backend only when the entrypoint explicitly sets `raw_fallback` to a TCP router and the fallback source policy allows the peer.
108
109```toml
110[entrypoints.edge]
111address = ":443"
112protocol = "smart_tcp"
113raw_fallback = "ssh-fallback"
114raw_fallback_trusted_cidrs = ["100.64.0.0/10"]
115```
116
117Use `raw_fallback_trusted_cidrs` for private raw fallback. Public raw fallback requires an explicit acknowledgement:
118
119```toml
120[entrypoints.edge]
121address = ":443"
122protocol = "smart_tcp"
123raw_fallback = "public-raw"
124allow_public_raw_fallback = true
125```
126
127Security behavior:
128
129- PROXY protocol v1 and v2 prefixes are rejected; Gordon does not trust or parse PROXY headers on smart TCP entrypoints.
130- `trusted_cidrs` and `raw_fallback_trusted_cidrs` use the peer socket IP, not `X-Forwarded-For`.
131- HTTP-looking or TLS-looking malformed traffic is rejected and never bypasses to raw fallback.
132- Entry-point-wide `trusted_cidrs` applies before sniffing to all protocols; raw fallback has its own narrower policy for private fallback exposure.
133
134## Docker network isolation
135
136Per-app Docker networks are enabled by default. To block direct external egress from isolated networks, opt into Docker internal networks:
137
138```toml
139[network_isolation]
140enabled = true
141internal = true
142```
143
144`internal = false` remains the compatibility default because some applications need direct egress during startup.
145
146Every app container joins an incarnation-owned private network derived from the app's internal UUID; shared memberships come only from explicit `[[network.shared]]` declarations, and memory/CPU/PID limits apply on every create and recovery path.
147
148## Readiness helper containers
149
150Probing an internal HTTP port uses one bounded helper container per readiness attempt, removed immediately after. Gordon force-removes the helper under an independent cleanup context, including on failure and timeout. There is no operator configuration for the helper; its image and limits are fixed.
151
152- The helper image is digest-pinned and multi-arch: `alpine@sha256:d9e853e87e55526f6b2917df91a2115c36dd7c696a35be12163d44e6e2a4b6bc`. The image must be pre-provisioned and available offline on the target host, because the probe never pulls.
153- The helper attaches only to the target app's private network. It never joins a shared network or a host network.
154- It never publishes a host port and has no mounts, volumes, secrets, environment, or runtime socket.
155- It runs non-root (uid/gid `65534`) with a read-only root filesystem, all capabilities dropped, `no-new-privileges`, and bounded CPU, memory, and PIDs.
156- It targets the exact inspected container IP on that network, never a service alias, and revalidates the container's execution start before trusting the result; a restarted generation is discarded.
157- A target that listens only on its own loopback fails readiness.
158- Helpers left behind by an abrupt daemon or host stop are reclaimed by the next probe once they are older than ten minutes; a live session is never touched.
159
160## Registry exposure
161
162- With `auth.enabled = true`, registry requests are authenticated and repository-scoped; the public proxy forwards registry domains to the internal registry.
163- With `auth.enabled = false`, the registry is local-only: the public proxy refuses registry-domain requests, and the registry handler accepts only direct loopback connections carrying the instance credentials.
164- Registry requests are parsed once into a validated operation used by both authorization and dispatch, so the repository a token is checked against is exactly the repository served.
165- Blob and upload access is repository-scoped: an upload UUID is usable only by the repository that started it, and a blob is served only to a repository that completed an upload of that digest.
166
167## Container runtime profile
168
169```toml
170[containers]
171security_profile = "compat" # or "strict"
172```
173
174- `compat` preserves broad image compatibility while retaining `no-new-privileges` and default capability restrictions.
175- `strict` enables a read-only root filesystem, drops all capabilities, and only adds `NET_BIND_SERVICE`.
176
177Use `strict` for images designed to write only to mounted volumes and run without extra Linux capabilities.
178
179## Related
180
181- [Auth](./auth.md)
182- [Images](./images.md)
183- [Network Isolation](./network-isolation.md)
184- [Volumes](./volumes.md)
185- [Reference](./reference.md)