2 Commits

Author	SHA1	Message	Date
YoungReckless4	56ac8d0fc7	feat(amneziawg): add opt-in Linux kernel datapath (#6877) ... * feat(amneziawg): add opt-in Linux kernel datapath Add the kernel backend as an explicit per-inbound option while keeping the embedded backend as the default. Include host preflight, UI/API support, the bundled installer CLI, lifecycle and accounting integration, regression tests, and upstream-facing documentation. Kernel mode is NAT-only and rejects unsupported routing, IPv6, forwarding, and pinned-listen configurations. Local saves check the local host; remote nodes check their own host during apply. Preserve adopted interface accounting across panel restarts, retain manager state after failed teardown so the job can retry, prevent embedded startup while a kernel interface remains live, and prune collector state for removed peers. Tested the restart accounting path on Debian 13 with XanMod 7.1.11 on x86_64; Docker and cross-kernel coverage are not claimed. * fix(amneziawg): align readiness and endpoint reporting Expire endpoint sightings using the latest handshake and sweep stale addresses for the polled interface. Keep the readiness API schema, generated client types, and ready-host empty blocker list consistent; watch the nested form field and defer local readiness checks for remote targets. Document safe rollback before using a build without kernel support. Regression evidence: the idle endpoint test failed before the collector change because the old address remained visible after 31 minutes; endpoint tests pass after the fix. The form component suite passes 3 tests; restoring the invalid array watch made its off-mode severity test fail, while the corrected dot-path watch passes. Targeted Go regressions across amneziawgkern, job, and service pass. * style(amneziawg): format readiness test * test: reproduce kernel AWG client validation gaps * test(service): cover kernel AWG shared client fanout * fix(amneziawg): validate kernel client fanout and config values * test(amneziawgkern): cover owned interface lifecycle recovery * fix(amneziawgkern): recover owned interfaces on restart * test(amneziawgkern): retain visibility after recovered cleanup failure * fix(amneziawgkern): report recovered interfaces while cleanup retries * style(amneziawgkern): satisfy static analysis * test(amneziawgkern): cover long rendered peer comments * fix(amneziawgkern): accept long generated config comments * test(amneziawgkern): refuse foreign occupied interfaces Capture the current Ensure ordering failure with real temporary config paths and manager side-effect boundaries. * test(amneziawgkern): cover stale interface collisions Extend the Ensure regression across absent links, mismatched live interfaces, cached manager state, and Remove teardown. * fix(amneziawgkern): verify interface ownership before mutation Refuse to replace or tear down links unless the canonical config is a valid panel-owned file and its key matches the live interface. Preserve normal adoption and key rotation, and document migration of legacy unmarked configs. Validation: go test ./internal/amneziawgkern -count=1; go test -race ./internal/amneziawgkern -count=1; golangci-lint run ./internal/amneziawgkern/...; git diff --check. * build: update Go toolchain to 1.27.3 * test(amneziawg): reproduce unsafe key rotation recovery Keep the previous owned config authoritative until teardown succeeds, and require fresh managers to rebuild structural key changes. These regressions expose the stuck live-key/disk-key mismatch after teardown failure. * fix(amneziawg): recover interrupted key rotations Keep the old owned config on disk until a structural rebuild has confirmed the previous interface is down. Fresh managers only adopt when the live config and desired structural identity match, so retries and removals retain the key needed to prove ownership. * fix(scanner): align Go image with module version * fix(docs): pin pnpm release with bootstrap fix pnpm 12.9.1 restores a Node launcher before its postinstall replaces the target with ELF; pnpm 11 then runs that binary as JavaScript. Pin 12.10.0, which includes the shim fix, so the docs CI bootstrap can switch versions safely. * fix(xray): build core from pinned module The configured Xray release asset tag is unpublished, and the latest published binary lacks GetUsersStats used by the panel. Build the packaged core from the exact Xray module version already selected in go.mod for Docker and release targets. * fix(tests): stabilize local network fixtures Keep the TUIC stream fragmentation check on a working UDP loopback path, allocate the AmneziaWG fixture ports dynamically, and wait for a real Xray stats response before starting API mutation assertions. Recreate failed readiness channels so an early connection cannot stall the probe. Verified host and offline parallel Go suites, repeated occupied-port and UDP tests, race/shuffle runs, and golangci-lint. Production behavior and dependency versions are unchanged. * fix(i18n): add the kernel-mode AmneziaWG keys to sk-SK main gained the sk-SK locale after this branch added its kernel-mode keys to the other 13, so i18n-dead-keys.test.ts failed on the merged tree. * refactor(amneziawgkern): apply kernel interfaces with direct commands, no awg-quick awg-quick evaluates PostUp/PostDown through a root shell, so every value the panel rendered into its config was one validation slip away from root command execution. The interface is now built with direct ip/awg/nft execs from a setconf-only file: link add, awg setconf, address, mtu, an atomically replaced nftables table per inbound, and the IPv4 forwarding sysctl written directly. - Panel-owned names and paths: xui-awg<id> interfaces, xui_awg<id> NAT tables and configs under the panel's data dir, so nothing collides with an operator's own awg0/awg1 awg-quick tunnels in /etc/amnezia/amneziawg. - FORWARD accepts (comment 3x-ui:xui-awg<id>) when iptables is present: Docker and ufw set the FORWARD policy to DROP, which silently black-holed the tunnel. - A full panel stop records the final counters and tears the interfaces down, like the embedded datapath; a stopped panel enforces no quota or expiry. Crash survivors are still adopted without a rebuild. - The job reads counters before reconciling, and from every managed interface, so a client disabled for quota, or a removed inbound, is charged for its last interval instead of losing it with the dropped peer. - An unchanged reconcile tick forks no process (it probed the live key with `awg show` every 10s); readiness no longer loads the module from a GET. Verified on the real module (WSL2 6.6 kernel, v3.1.20260906): create, peer and egress changes in place, SIGKILL adoption with the same ifindex, SIGTERM teardown leaving no link, table, FORWARD rule or config behind. * fix(amneziawg): gate kernel host readiness only when a save starts the datapath Every save of a kernel-mode inbound re-ran the host preflight, so a host that lost its module (a kernel upgrade without a DKMS rebuild) refused even disabling the inbound or switching it back to embedded mode. Host blockers now apply only when the save starts the kernel datapath: a new inbound, a switch from embedded, or enabling a disabled one. Feature blockers (IPv6, forwarded ports, pinned listen) still apply to every save. * feat(nodes): refuse kernel-mode AmneziaWG pushes to nodes that predate it A node built before kernel mode drops the unknown kernelMode field and serves the inbound on the embedded datapath while the master's form shows kernel mode on. Nodes now advertise awg-kernel in X-3x-Node-Caps, and Remote refuses to push a kernel-mode inbound to a node that has not advertised it, learning the caps with an inbound list first when this Remote has not heard from it. * fix(amneziawg): pin the kernel module installer to verified commits The installer ran as root and built whatever upstream tag sorted newest at install time. It now clones fixed tags (module v3.1.20260906, tools v3.1.20260812) and refuses unless each resolves to its recorded commit. The udp_tunnel perl patch is gone: those releases carry the kernel 7.1.5 compat upstream, so the patch never matched. * fix(i18n): translate the kernel-mode AmneziaWG strings All thirteen non-English locales carried the English text verbatim. * fix(amneziawg): skip the kernel traffic flush when no kernel interface runs Shutdown listed the desired AmneziaWG inbounds from the database even on a panel with no kernel interface, an unneeded query on every stop. A server stopped before its database was opened panicked on it, which turned TestServerStopWhenXrayNotRunningSucceeds red whenever it ran first. --------- Co-authored-by: MHSanaei <ho3ein.sanaei@gmail.com>	2026-10-25 13:24:50 +01:00
Sanaei	37c5e0bfd2	feat(node): node hardening — mTLS, hashed+zstd reconcile transport, per-node net metrics (#5382) ... * fix(api-docs): document clientIpsByGuid route Restores a green `go test ./...` baseline: TestAPIRoutesDocumented flagged POST /panel/api/clients/clientIpsByGuid (added in 9385b6c6) as undocumented in endpoints.ts. * test(node): characterize current node TLS + API auth behavior Phase 0 regression net for the mTLS work. These pass on unchanged production code and lock the pre-mTLS contracts so later phases can be proven additive: - tlsConfigForNode: skip -> InsecureSkipVerify (no VerifyConnection); pin -> VerifyConnection installed. - checkAPIAuth: bearer match -> Next + api_authed; unauthenticated -> 401 (XHR) / 404; valid session -> Next. - panel HTTPS listener with no ClientAuth accepts a client that presents no client certificate (the browsers-keep-working invariant). * feat(crypto): node-auth CA + client-cert minting (TDD) Stdlib-only ECDSA P-256 helpers for the node mTLS work: - GenerateNodeCA: self-signed CA (IsCA, CertSign, path len 0) - IssueClientCert: client-auth leaf (ExtKeyUsageClientAuth) signed by CA - LoadCAFromPEM: parse a CA cert+key for issuing / trust-pool building Tests assert the contract (leaf verifies against the issuing CA with ExtKeyUsageClientAuth), seen failing on the assertion before impl. * feat(node): lazy node mTLS CA + client cert in settings (TDD) SettingService gains opt-in mTLS material, all stored as Setting rows with empty defaults and kept out of entity.AllSetting (so private keys never reach the settings UI/export): - EnsureNodeMtlsCA: mint+persist the node-auth CA once, reuse thereafter - EnsureMasterClientCert: issue the master client cert from the CA, idempotent - NodeMtlsClientCAPool: ClientCAs trust pool for the listener; nil when unconfigured so the no-mTLS path is unchanged Tests assert idempotency and that the client cert verifies against the CA for client auth; seen failing on the assertion before impl. * feat(node): mtls client TLS config + master-cert provider (TDD) tlsConfigForNode gains an 'mtls' branch that presents the master client certificate and verifies the node server against system roots (no InsecureSkipVerify, no custom RootCAs). The cert is supplied via an injected MasterClientCertProvider so runtime need not import service; it fails closed when unconfigured. skip/pin contracts unchanged. * feat(node): allow tokenless mtls nodes in remote do() (TDD) mtls nodes authenticate with a client certificate, so the bearer token becomes optional for them: do() no longer rejects an empty ApiToken when TlsVerifyMode is mtls, and the Authorization header is omitted when no token is set. Every other mode still requires a token (regression kept). * feat(node): authenticate verified client certs in checkAPIAuth (TDD) A completed mTLS handshake (non-empty r.TLS.VerifiedChains) now authenticates an API request, equivalent to a valid bearer token, and sets api_authed so the CSRF middleware lets cert-authed mutations through. Bearer/session/reject paths unchanged. The accept-path assert was mutation-checked (guard flipped -> test red -> reverted). * feat(node): opt-in mTLS on the panel listener (TDD; mutation-checked) web.go now applies VerifyClientCertIfGiven + ClientCAs to the HTTPS listener when a node trust CA is configured, and wires the master client cert provider for outbound mtls calls. With no CA the listener is byte-identical to before (browsers unaffected). applyNodeMtls is covered end-to-end: no-cert client handshakes (browsers keep working), a CA-signed client cert verifies, a foreign-CA cert is rejected at the handshake. Mutation-checked: - RequireAndVerifyClientCert -> no-cert client rejected (red) -> reverted - drop ClientCAs -> master cert no longer trusted (red) -> reverted * feat(node): accept mtls verify-mode + CA reveal endpoint (TDD) - model.Node.TlsVerifyMode validator now accepts 'mtls' - normalize() preserves mtls and requires the node scheme to be https (fail closed), instead of clamping mtls back to verify - NodeService.NodeMtlsCaCert + POST /panel/api/nodes/mtls/ca return this panel's node-auth CA cert (public) to paste into a node, minting the CA + master client cert on first call - endpoints.ts documents the new route (doc-sync test) No model column added (enum is a string), so no migration/codegen. * feat(node): node mTLS UI + trust-CA setter (TDD) Backend: - NodeService.SetNodeMtlsTrustCA + POST /panel/api/nodes/mtls/trustCA store the CA this panel trusts for incoming node-API client certs (validates PEM, empty clears); applied on next restart - endpoints.ts + regenerated openapi.json document both mtls routes Frontend: - node form: 'mtls' TLS-verify option + setup hint (zod enum updated) - Nodes page 'Node mTLS' card: copy this panel's CA, and paste/save the trusted parent CA - en-US i18n keys (other locales fall back to en-US) Gates green: go build (native+windows), vet, go test ./...; frontend typecheck, lint, vitest (541). * style(node): gofmt web_mtls_test doc comment * feat(node): hashed+zstd reconcile transport (TDD, negotiated, mixed-version safe) Adds an integrity + compression envelope to node config pushes: - internal/util/wirecodec: shared zstd codec (bomb-capped decode) + SHA-256 hashing + the header/capability constants - Remote.do(): always attaches X-Config-Sha256 of the uncompressed body; zstd-compresses only when the node advertised support (learned from its X-3x-Node-Caps response header) and the body is >=1KiB - ConfigEnvelopeMiddleware on /panel/api: advertises the cap, decompresses and verifies the hash (handler not invoked on mismatch) before binding Mixed-version safe: old nodes never advertise the cap -> plain bodies; the hash header is verify-if-present so any panel/node mix interoperates (existing reconcile tests stay green). klauspost/compress promoted to a direct dep. Hash-mismatch reject was mutation-checked (compare defeated -> test red -> reverted). * feat(node): per-node network throughput metrics (TDD) The node status response already carries gopsutil netIO.up/down (summed non-virtual interfaces), so no node-side change is needed: - probe() parses netIO.up/down into HeartbeatPatch.NetUp/NetDown - Node gains net_up/net_down columns (AutoMigrate); UpdateHeartbeat persists them and appends netUp/netDown to the per-node metric history - NodeMetricKeys whitelists netUp/netDown so the history endpoint serves them - NodeHistoryPanel renders Net Up/Down sparklines (KB/s, no 0-100 clamp) - regenerated frontend types + openapi.json for the new Node fields * feat(node): move node mTLS controls into a toolbar button + modal The Node mTLS panel was an always-visible card cluttering the nodes page. Replace it with a 'Node mTLS' button beside 'Add node' that opens a modal with the same copy-CA + trusted-parent-CA controls; the modal closes on a successful save. No backend/i18n changes. * i18n(node): translate mTLS + net-metrics keys for all locales Adds the node mTLS strings (tlsMtls, mtlsFormHint, mtls.* dialog + the saveMtls toast) and the netUp/netDown chart labels to all 12 non-English catalogs (ar, es, fa, id, ja, pt, ru, tr, uk, vi, zh-CN, zh-TW), matching each catalog's existing terminology. Technical tokens (mTLS/TLS/CA/API/ KB/s) kept verbatim. * fix(node): address Copilot review on node-hardening PR - setting_mtls: fail closed on a half-present CA/master-cert pair instead of silently regenerating (which would rotate the CA and break fleet trust). - config_envelope: reject non-zstd Content-Encoding on the envelope path rather than hashing/forwarding a still-encoded body to the handler. - node mTLS: support tokenless mTLS end-to-end — apiToken is now required_unless tlsVerifyMode=mtls (model) with matching conditional validation in NodeFormSchema, so the runtime allowance is actually reachable. - NodesPage: add a catch block to onSaveTrustCa so save failures surface.	2026-06-16 12:19:33 +02:00