Files
LocalAI/pkg/natsauth/permissions_coverage_test.go
Richard Palethorpe 3a932a9803 feat(distributed): Add NATS JWT authentication and TLS/mTLS options (#10159)
* feat(distributed): NATS JWT auth, TLS/mTLS options, and e2e coverage

Mint per-node NATS user JWTs at registration when LOCALAI_NATS_ACCOUNT_SEED
is set, and connect workers with scoped credentials from the register response.
Add optional LOCALAI_NATS_TLS_CA/CERT/KEY for private CA and mTLS alongside
tls:// URLs, plus test-e2e-distributed and NatsJWT container e2e specs.

Document JWT setup (nats-auth-setup.sh) and TLS env vars in distributed-mode.

Assisted-by: Grok:grok grok-build
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* fix(distributed): correct NATS JWT scoping and harden client auth

The JWT-auth path added in 46467cc7 had several gaps that fail silently
under LOCALAI_NATS_REQUIRE_AUTH:

- Agent-worker minted JWTs did not allow the subjects the agent worker
  actually subscribes to (jobs.mcp-ci.new and nodes.<id>.backend.stop),
  so MCP-CI jobs and backend-stop session cleanup were silently dropped.
  Scope the agent permission set to those subjects.
- NATS subscription permission violations were swallowed (Subscribe
  returned a live-but-dead subscription). Confirm subscriptions with a
  server round-trip so a denial surfaces synchronously, and log async
  permission errors.
- The backend worker connected anonymously when given a JWT without its
  paired seed; reject the unpaired credential instead.
- The documented service-user permissions in nats-auth-setup.sh omitted
  prefixcache.>, which the frontend publishes and subscribes; add it.

Also: add a credential-provider hook to the messaging client (consumed by
the follow-up credential-lifecycle change), drop the always-nil error from
NatsMessagingOptions, run go mod tidy (jwt/v2 and nkeys are now direct),
and gofmt the feature's files.

Tests: an agent-JWT e2e spec that connects to the enforcing NATS server
and exercises every subscription the agent worker makes, plus permission
allow-list coverage unit tests.

Assisted-by: Claude:claude-opus-4-8 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* feat(distributed): acquire and auto-refresh worker NATS credentials

Workers fetched NATS credentials once at startup, which broke two cases
under JWT auth: a worker that registered while still pending admin
approval never received a minted JWT (it connected unauthenticated and
gave up), and a long-running worker's 24h JWT expired with no way to renew
it.

Introduce workerregistry.NATSCredentialManager, built on idempotent
re-registration (the frontend preserves the node row and mints a fresh JWT
each call):

- Acquire re-registers through admin approval until the node is approved
  and credentials are minted (or returns the first success when auth is
  not required, preserving anonymous-NATS behavior).
- RefreshLoop re-registers before the JWT expires (~75% of its lifetime),
  updating the credentials served to the connection.
- Both are bounded (default 100 attempts / consecutive failures) and
  return an error on exhaustion, so an unapprovable or unrenewable worker
  exits non-zero and surfaces the problem instead of hanging or drifting
  toward an expired credential.

The messaging client gains WithUserJWTProvider, fetching credentials on
each (re)connect so the connection transparently adopts a refreshed JWT
when the server expires the old one. RegisterFull exposes the approval
status and full response; Register delegates to it.

Both the backend worker and the agent worker are wired to this: explicit
env credentials are used as-is, minted credentials are acquired-with-wait
and refreshed, and a permanent refresh failure shuts the worker down so it
restarts and re-acquires.

Tests cover Acquire (wait-through-pending, bounded give-up, context
cancel), RefreshLoop (refresh-before-expiry, bounded failure, no-expiry
exit) and jwtExpiry decoding. Docs updated in distributed-mode.md.

Assisted-by: Claude:claude-opus-4-8 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

---------

Signed-off-by: Richard Palethorpe <io@richiejp.com>
2026-06-03 19:43:56 +02:00

135 lines
4.7 KiB
Go

package natsauth_test
import (
"os"
"regexp"
"strings"
"github.com/mudler/LocalAI/core/services/messaging"
"github.com/mudler/LocalAI/pkg/natsauth"
. "github.com/onsi/ginkgo/v2"
. "github.com/onsi/gomega"
)
// subjectMatches implements NATS subject-token matching: "*" matches exactly one
// token and ">" matches one or more trailing tokens. It lets these tests assert
// that a permission allow-list (which uses wildcards) actually covers a concrete
// subject a component publishes/subscribes — the same check the NATS server makes.
func subjectMatches(pattern, subject string) bool {
p := strings.Split(pattern, ".")
s := strings.Split(subject, ".")
for i, tok := range p {
if tok == ">" {
return i < len(s) // ">" must match at least one remaining token
}
if i >= len(s) {
return false
}
if tok != "*" && tok != s[i] {
return false
}
}
return len(p) == len(s)
}
func anyAllows(allow []string, subject string) bool {
for _, p := range allow {
if subjectMatches(p, subject) {
return true
}
}
return false
}
var _ = Describe("WorkerPermissions subject coverage", func() {
// A node ID containing NATS-reserved characters exercises the (duplicated)
// sanitizer in pkg/natsauth against the canonical one in core/services/messaging.
// If the two ever diverge, the minted prefix stops matching the real subject
// and these assertions fail — guarding the copy noted in the review.
const nodeID = "host.a 1*b"
Context("backend worker", func() {
pub, sub := natsauth.WorkerPermissions(nodeID, "backend")
// Every subject core/services/worker/{lifecycle,file_staging}.go subscribes to.
subscribed := []string{
messaging.SubjectNodeBackendInstall(nodeID),
messaging.SubjectNodeBackendUpgrade(nodeID),
messaging.SubjectNodeBackendStop(nodeID),
messaging.SubjectNodeBackendDelete(nodeID),
messaging.SubjectNodeBackendList(nodeID),
messaging.SubjectNodeModelUnload(nodeID),
messaging.SubjectNodeModelDelete(nodeID),
messaging.SubjectNodeStop(nodeID),
messaging.SubjectNodeFilesEnsure(nodeID),
messaging.SubjectNodeFilesStage(nodeID),
messaging.SubjectNodeFilesTemp(nodeID),
messaging.SubjectNodeFilesListDir(nodeID),
}
for _, subject := range subscribed {
It("allows subscribing to "+subject, func() {
Expect(anyAllows(sub, subject)).To(BeTrue(),
"backend JWT sub allow-list %v does not cover %s", sub, subject)
})
}
It("allows publishing backend.install progress", func() {
subject := messaging.SubjectNodeBackendInstallProgress(nodeID, "op-123")
Expect(anyAllows(pub, subject)).To(BeTrue(),
"backend JWT pub allow-list %v does not cover %s", pub, subject)
})
})
Context("agent worker", func() {
// node_type "agent"; subjects from core/cli/agent_worker.go.
pub, sub := natsauth.WorkerPermissions(nodeID, "agent")
_ = pub
subscribed := []string{
messaging.SubjectAgentExecute, // dispatcher (default --agent-subject)
messaging.SubjectMCPToolExecute, // QueueSubscribeReply
messaging.SubjectMCPDiscovery, // QueueSubscribeReply
messaging.SubjectMCPCIJobsNew, // QueueSubscribe — jobs.mcp-ci.new
messaging.SubjectNodeBackendStop(nodeID), // Subscribe — MCP session cleanup
}
for _, subject := range subscribed {
It("allows subscribing to "+subject, func() {
Expect(anyAllows(sub, subject)).To(BeTrue(),
"agent JWT sub allow-list %v does not cover %s — the agent worker subscribes to it", sub, subject)
})
}
})
})
var allowPubRe = regexp.MustCompile(`--allow-pub "([^"]*)"`)
var _ = Describe("Documented NATS service-user permissions", func() {
// scripts/nats-auth-setup.sh ships the recommended service (frontend) JWT
// permissions. They must cover every subject the frontend actually publishes,
// or prefix-cache sync (and friends) break once LOCALAI_NATS_REQUIRE_AUTH is on.
const scriptPath = "../../scripts/nats-auth-setup.sh"
// Representative subjects the frontend publishes on the control plane.
// prefixcache.* is emitted by prefixcache.Sync in core/application/distributed.go.
frontendPublishes := []string{
messaging.SubjectPrefixCacheObserve,
messaging.SubjectPrefixCacheInvalidate,
messaging.SubjectNodeBackendInstall("node-1"),
messaging.SubjectGalleryProgress("op-1"),
}
It("cover every subject the frontend publishes", func() {
raw, err := os.ReadFile(scriptPath)
Expect(err).ToNot(HaveOccurred(), "cannot read %s", scriptPath)
m := allowPubRe.FindStringSubmatch(string(raw))
Expect(m).To(HaveLen(2), "no --allow-pub list found in %s", scriptPath)
allow := strings.Split(m[1], ",")
for _, subject := range frontendPublishes {
Expect(anyAllows(allow, subject)).To(BeTrue(),
"service-user --allow-pub %v does not cover %s (frontend publishes it)", allow, subject)
}
})
})