mirror of
https://github.com/jmcorgan/fips.git
synced 2026-10-30 07:37:36 +00:00
Johnathan Corgan
1cc069f1a2
The TUN adapter, the .fips DNS responder, ICMPv6 generation and TCP MSS clamping are the host-side IPv6 plane, but the code was split between src/upper and inline bodies in the node's session handler and lifecycle. Gather it into one module, src/ipv6tun, and reduce its calls into the rest of the crate to a small explicit set. This prepares the TUN adapter to run later as a separate daemon over the native API, and gives embedders one place to find the host-side surface. Behaviour is unchanged apart from log target names. - Rename src/upper to src/ipv6tun. A `pub use ipv6tun as upper;` alias keeps every crate::upper:: and fips::upper:: path resolving, so no consumer has to change. - Move hosts.rs whole to src/hosts.rs, a top-level public module: the hosts file also serves peer display names, the peer ACL and fipsctl, so it is not host-side only. ipv6tun re-exports it for the old path. - Move the DNS socket helpers (dual-stack bind, IPV6_RECVPKTINFO, interface index lookup) from Node into ipv6tun::dns, unchanged. Node::mesh_ifindex, which reads the live TUN device name, becomes Handles::mesh_ifindex. - Move ICMPv6 Destination Unreachable and Packet Too Big sending into ipv6tun::icmp behind IcmpContext, which borrows the TUN channel, our address and the Packet Too Big rate limiter for one use. Packet Too Big is still rate limited and Destination Unreachable still is not. The discovery lookup timeout hands its queued packets over as one no-route report. - Split handle_tun_outbound. The host-side half, in ipv6tun::outbound, validates the packet, makes both Packet Too Big decisions and sends the ICMPv6 replies, reaching the mesh through a small Mesh trait Node implements. The mesh-side half, Node::send_outbound, stays with the pending queue. The checks run in the same order with the same thresholds, and Node::handle_tun_outbound remains the entry point. - Move the TUN and DNS child start and stop bodies into ipv6tun::lifecycle, and gather their nine supervisor fields and the node's TUN device name into one Handles struct the supervisor holds. The supervisor arms, their order and the child-exit reporting are unchanged. A TUN still counts as up when it has a device name, so an app-owned TUN produces no TUN teardown, and DNS counts as up while its task runs. The node passes a new peer-alias base to the running responder through Handles::publish_aliases. Node::tun_name, tun_tx, dns_local_addr and enable_app_owned_tun keep their behaviour; tests install a TUN sender through a test-only Node::install_tun. Tracing targets follow module paths, so lines from the moved code now log under fips::ipv6tun::* and fips::hosts instead of fips::upper::*, fips::node::lifecycle and fips::node::handlers::session. Update the RUST_LOG example in the MTU diagnosis guide and the test harness filters that relied on the old targets, and note the rename in the changelog.
173 lines
5.9 KiB
Markdown
173 lines
5.9 KiB
Markdown
# ACL Allowlist Test
|
|
|
|
Six Docker nodes use per-node ACL files mounted at the hardcoded runtime paths:
|
|
|
|
- `node-a` and `node-b` carry the insider allowlist (`node-a`, `node-b`, `node-e`, `node-f`)
|
|
- `node-c` and `node-d` each carry a broad allowlist containing every node alias
|
|
- `node-e` and `node-f` do not mount any ACL files locally
|
|
- every node gets a generated `/etc/fips/hosts` with aliases for `node-a` through `node-f`
|
|
|
|
This lets us test three different node behaviors at once:
|
|
|
|
- insiders (`a`, `b`) explicitly allow `a`, `b`, `e`, and `f`
|
|
- outsiders (`c`, `d`) allow everyone locally, but still cannot join because insiders reject them
|
|
- allowed remotes (`e`, `f`) rely on the insider ACLs and do not need local ACL files
|
|
|
|
## Test Identities
|
|
|
|
Allowed:
|
|
|
|
- `node-a`
|
|
- `npub1sjlh2c3x9w7kjsqg2ay080n2lff2uvt325vpan33ke34rn8l5jcqawh57m`
|
|
- `0102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f20`
|
|
- `node-b`
|
|
- `npub1tdwa4vjrjl33pcjdpf2t4p027nl86xrx24g4d3avg4vwvayr3g8qhd84le`
|
|
- `b102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1fb0`
|
|
|
|
Denied:
|
|
|
|
- `node-c`
|
|
- `npub1cld9yay0u24davpu6c35l4vldrhzvaq66pcqtg9a0j2cnjrn9rtsxx2pe6`
|
|
- `c102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1fc0`
|
|
- `node-d`
|
|
- `npub1n9lpnv0592cc2ps6nm0ca3qls642vx7yjsv35rkxqzj2vgds52sqgpverl`
|
|
- `d102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1fd0`
|
|
|
|
Additional allowed:
|
|
|
|
- `node-e`
|
|
- `npub1x5z9rwzzm26q9verutx4aajhf2zw2pyp34c6whhde2zduxqav40qgq36l6`
|
|
- `nsec1egyrmekfw3u4l88v8zhrak9uht503s2kvn9v49tqgp6c5l2yuxgsv386l0`
|
|
- `node-f`
|
|
- `npub1ytrut7gjncn2zfnhn56c0zgftf0w6p99gf6fu8j73hzw5603zglqc9av6c`
|
|
- `nsec1afh3nysthqh47awpdewcw59wvvp499f8dvlyclmnv4gvpxdk56dsa6eqsn`
|
|
|
|
The generated `fips.key` fixtures use a mix of bare hex and `nsec1...` values.
|
|
FIPS accepts either format in key files.
|
|
|
|
## Run
|
|
|
|
Build the Linux binaries and test image:
|
|
|
|
```bash
|
|
./testing/scripts/build.sh --no-docker
|
|
```
|
|
|
|
Start the ACL test mesh:
|
|
|
|
```bash
|
|
./testing/acl-allowlist/generate-configs.sh
|
|
docker compose -f testing/acl-allowlist/docker-compose.yml up -d --build
|
|
```
|
|
|
|
Or run the full integration check:
|
|
|
|
```bash
|
|
./testing/acl-allowlist/test.sh
|
|
```
|
|
|
|
`test.sh` regenerates the ACL fixtures automatically before starting Docker.
|
|
The generated ACL files use alias names, and the generated hosts file makes
|
|
those aliases resolvable at runtime.
|
|
|
|
The ACL harness pins the expected test entrypoint explicitly so it does not
|
|
accidentally reuse an older `fips-test:latest` image with a different startup
|
|
script.
|
|
|
|
Docker service/container/hostname identifiers in this harness intentionally use
|
|
`service-*`, `fips-acl-container-*`, and `host-*` names so they do not collide with the logical FIPS aliases
|
|
`node-a` through `node-f`. For data-plane checks and operator examples, use the
|
|
explicit FIPS names such as `node-a.fips` and `node-d.fips`.
|
|
|
|
The bridge network requests no subnet, so docker assigns one from its own
|
|
address pool and two concurrent runs of this harness never contend for a fixed
|
|
range. Consequently no node's IPv4 address is known before startup, and the
|
|
generated peer stanzas address each other by docker hostname (`host-a` …
|
|
`host-f`), resolved through the container's dnsmasq to docker's embedded DNS.
|
|
|
|
ACL paths are fixed in this branch:
|
|
|
|
- `/etc/fips/peers.allow`
|
|
- `/etc/fips/peers.deny`
|
|
|
|
Mounted ACL files in this harness:
|
|
|
|
- `node-a` and `node-b`: insider allowlist plus `ALL` deny fallback
|
|
- `node-c` and `node-d`: broad local allowlist used by outsider nodes trying to blend in
|
|
- `node-e` and `node-f`: no ACL files mounted
|
|
- all nodes: `/etc/fips/hosts` aliases for `node-a` through `node-f`
|
|
|
|
Generated fixture location:
|
|
|
|
- `testing/acl-allowlist/generated-configs/`, or
|
|
`generated-configs<suffix>/` when `FIPS_CI_NAME_SUFFIX` is set, which is how
|
|
concurrent runs keep their fixtures apart
|
|
|
|
Inspect peer state:
|
|
|
|
```bash
|
|
docker exec fips-acl-container-a fipsctl show peers
|
|
docker exec fips-acl-container-b fipsctl show peers
|
|
docker exec fips-acl-container-c fipsctl show peers
|
|
docker exec fips-acl-container-d fipsctl show peers
|
|
docker exec fips-acl-container-e fipsctl show peers
|
|
docker exec fips-acl-container-f fipsctl show peers
|
|
```
|
|
|
|
Inspect the loaded ACL state directly:
|
|
|
|
```bash
|
|
docker exec fips-acl-container-a fipsctl acl show
|
|
```
|
|
|
|
Use explicit `.fips` names when checking reachability through the FIPS overlay:
|
|
|
|
```bash
|
|
docker exec fips-acl-container-a ping node-d.fips
|
|
docker exec fips-acl-container-a ping npub1n9lpnv0592cc2ps6nm0ca3qls642vx7yjsv35rkxqzj2vgds52sqgpverl.fips
|
|
```
|
|
|
|
The output shows both the original ACL file entries and the resolved
|
|
effective npub entries.
|
|
|
|
Expected:
|
|
|
|
- `node-a` sees `node-b`, `node-e`, and `node-f`
|
|
- `node-b` sees `node-a`
|
|
- `node-c` sees no peers
|
|
- `node-d` sees no peers
|
|
- `node-e` sees `node-a`
|
|
- `node-f` sees `node-a`
|
|
|
|
Visible rejection logs:
|
|
|
|
```bash
|
|
docker compose -f testing/acl-allowlist/docker-compose.yml logs -f service-a service-b service-c service-d service-e service-f
|
|
```
|
|
|
|
On startup, `node-c` and `node-d` immediately try their configured outbound
|
|
static connection to `node-a`. Their own ACLs permit that attempt, but the
|
|
insider ACL on `node-a` still rejects both peers. Because `node-a` also has
|
|
static peer stanzas for `node-c` and `node-d`, you may see both
|
|
`outbound_connect` and `inbound_handshake` rejection messages during startup.
|
|
The outsider-initiated path emits messages like:
|
|
|
|
```text
|
|
Rejected peer by ACL ... context=inbound_handshake decision=denylist match
|
|
```
|
|
|
|
Those messages are now emitted at debug level. This harness enables
|
|
`fips::node=debug` in `RUST_LOG` so the ACL rejection details stay visible in
|
|
test logs, and operators can temporarily raise log level the same way when
|
|
diagnosing ACL issues locally.
|
|
|
|
A later `ping6` from `node-c.fips` does not emit a new `inbound_handshake` message.
|
|
The ping uses the data-plane session path, and since no peer session to
|
|
`node-a.fips` was established, it just times out.
|
|
|
|
Stop and clean up:
|
|
|
|
```bash
|
|
docker compose -f testing/acl-allowlist/docker-compose.yml down
|
|
```
|