Files
seaweedfs/test/s3/iam/s3_sts_assume_role_test.go
T
Chris Lu e77f8ae204 fix(s3api): route STS GetFederationToken to STS handler (#9157) (#9167)
* fix(s3api): route STS GetFederationToken requests to STS handler (#9157)

The STS GetFederationToken handler was implemented but never reachable.
Three routing gaps sent requests to the S3/IAM path instead of STS:

- No explicit mux route for Action=GetFederationToken in the URL query
- iamMatcher did not exclude GetFederationToken, so authenticated POSTs
  with Action in the form body were matched and dispatched to IAM
- UnifiedPostHandler only dispatched AssumeRole* and GetCallerIdentity
  to STS, leaving GetFederationToken to fall through to DoActions and
  return NotImplemented

Add the missing route, the matcher exclusion, and the dispatch branch.

Also wire TestSTS, TestAssumeRoleWithWebIdentity, and TestServiceAccount
into the s3-iam-tests workflow as a new "sts" matrix entry. Before this
change, none of test/s3/iam/s3_sts_get_federation_token_test.go's four
test functions ran in CI, which is why this regression shipped.

* test(iam): make orphaned STS/service-account tests pass under auth-enabled CI

Follow-up to wiring STS tests into CI: fixes several pre-existing issues
that made the newly-included tests fail locally.

Server fixes:
- weed/s3api/s3api_sts.go: handleGetFederationToken no longer 500s when
  the caller is a legacy S3-config identity (not in the IAM user store).
  Previously any GetPoliciesForUser error short-circuited to InternalError,
  which hard-failed every SigV4 caller using keys from -s3.config.
- weed/s3api/s3api_embedded_iam.go: CreateServiceAccount now generates IDs
  in the sa:<parent>:<uuid> format required by
  credential.ValidateServiceAccountId. The old "sa-XXXXXXXX" format failed
  the persistence-layer regex and caused every CreateServiceAccount call
  to return 500 once a filer-backed credential store validated the ID.

Test helpers:
- test/s3/iam/s3_sts_assume_role_test.go: callSTSAPIWithSigV4 no longer
  sets req.Header["Host"]. aws-sdk-go v1 v4.Signer already signs Host from
  req.URL.Host, and a manual Host header made the signer emit host;host in
  SignedHeaders, producing SignatureDoesNotMatch. Updated missing_role_arn
  subtest to match the existing SeaweedFS behavior (user-context
  assumption).
- test/s3/iam/s3_service_account_test.go: callIAMAPI now SigV4-signs
  requests when STS_TEST_{ACCESS,SECRET}_KEY env vars are set. Unsigned
  IAM writes otherwise fall through to the STS fallback and return
  InvalidAction.

CI matrix:
- .github/workflows/s3-iam-tests.yml: skip
  TestServiceAccountLifecycle/use_service_account_credentials only. The
  rest of the service-account suite passes; that one subtest depends on a
  separate credential-reload issue where new ABIA keys briefly register
  into accessKeyIdent but aren't persisted to the filer, so they vanish
  on the next reload. Out of scope for the #9157 GetFederationToken fix.

* fix(credential): accept AWS IAM username chars in service-account IDs

Gemini review on #9167 pointed out that ServiceAccountIdPattern's
parent-user segment was more restrictive than an AWS IAM username:
`[A-Za-z0-9_-]` vs. IAM's `[\w+=,.@-]`. Realistic usernames with
`@`, `.`, `+`, `=`, or `,` (e.g. email-style principals) would fail
validation at the filer store even though the embedded IAM API
happily created them.

Broaden the regex to `[A-Za-z0-9_+=,.@-]` (matching the AWS IAM spec
at https://docs.aws.amazon.com/IAM/latest/APIReference/API_User.html)
and add a table-driven test that locks the expansion in.

* address PR review feedback on #9167

All five review items were valid; changes keyed to review bullets:

- weed/s3api/s3api_sts.go: handleGetFederationToken no longer swallows
  arbitrary policy-lookup failures. Only credential.ErrUserNotFound is
  treated leniently (the legacy-config SigV4 path); any other error now
  returns InternalError so we don't mint tokens with an incomplete
  policy set.
- weed/credential/grpc/grpc_identity.go: GetUser translates gRPC
  NotFound back to credential.ErrUserNotFound so errors.Is(...) above
  matches for gRPC-backed stores, not just memory/filer-direct.
- weed/s3api/s3api_embedded_iam.go: CreateServiceAccount now validates
  the generated saId against credential.ValidateServiceAccountId before
  returning. Surfaces a client 400 with the offending ID instead of the
  opaque 500 that used to bubble up from the persistence layer.
- weed/s3api/s3api_server_routing_test.go: seed a routing-test identity
  with a known AK/SK, sign TestRouting_GetFederationTokenAuthenticatedBody
  with aws-sdk-go v4.Signer so the request actually passes
  AuthSignatureOnly. Assert 503 ServiceUnavailable (from STSHandlers
  with no stsService) instead of just NotEqual(501) — 503 proves the
  dispatch reached STSHandlers.HandleSTSRequest.
- test/s3/iam/s3_service_account_test.go: callIAMAPI signs with
  service="iam" instead of "s3" (SeaweedFS verifies against whichever
  service the client signed with, but "iam" is semantically correct).
- weed/credential/validation_test.go: add positive rows for an
  uppercase parent (sa:ALICE:...) and a canonical hyphenated UUID
  suffix (sa:alice:123e4567-e89b-12d3-a456-426614174000).
2026-04-20 19:33:22 -07:00

366 lines
12 KiB
Go

package iam
import (
"encoding/xml"
"fmt"
"io"
"net/http"
"net/url"
"os"
"strings"
"testing"
"time"
"github.com/aws/aws-sdk-go/aws/credentials"
v4 "github.com/aws/aws-sdk-go/aws/signer/v4"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// AssumeRoleResponse represents the STS AssumeRole response
type AssumeRoleTestResponse struct {
XMLName xml.Name `xml:"AssumeRoleResponse"`
Result struct {
Credentials struct {
AccessKeyId string `xml:"AccessKeyId"`
SecretAccessKey string `xml:"SecretAccessKey"`
SessionToken string `xml:"SessionToken"`
Expiration string `xml:"Expiration"`
} `xml:"Credentials"`
AssumedRoleUser struct {
AssumedRoleId string `xml:"AssumedRoleId"`
Arn string `xml:"Arn"`
} `xml:"AssumedRoleUser"`
} `xml:"AssumeRoleResult"`
}
// TestSTSAssumeRoleValidation tests input validation for AssumeRole endpoint
func TestSTSAssumeRoleValidation(t *testing.T) {
if testing.Short() {
t.Skip("Skipping integration test in short mode")
}
if !isSTSEndpointRunning(t) {
t.Fatal("SeaweedFS STS endpoint is not running at", TestSTSEndpoint, "- please run 'make setup-all-tests' first")
}
// Check if AssumeRole is implemented by making a test call
if !isAssumeRoleImplemented(t) {
t.Fatal("AssumeRole action is not implemented in the running server - please rebuild weed binary with new code and restart the server")
}
t.Run("missing_role_arn_defaults_to_caller_identity", func(t *testing.T) {
// SeaweedFS intentionally allows AssumeRole without RoleArn to support
// S3-compatible clients that omit it. In that case the session is tied
// to the caller's own identity (User Context assumption). See
// handleAssumeRole in weed/s3api/s3api_sts.go.
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleSessionName": {"test-session"},
// RoleArn is missing on purpose
}, "test-access-key", "test-secret-key")
require.NoError(t, err)
defer resp.Body.Close()
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
assert.Equal(t, http.StatusOK, resp.StatusCode,
"Should succeed without RoleArn (user-context assumption): %s", string(body))
var assumeResp AssumeRoleTestResponse
require.NoError(t, xml.Unmarshal(body, &assumeResp),
"Failed to parse response: %s", string(body))
assert.NotEmpty(t, assumeResp.Result.Credentials.AccessKeyId,
"AccessKeyId should be issued for user-context assumption")
})
t.Run("missing_role_session_name", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/test-role"},
// RoleSessionName is missing
}, "test-access-key", "test-secret-key")
require.NoError(t, err)
defer resp.Body.Close()
assert.NotEqual(t, http.StatusOK, resp.StatusCode,
"Should fail without RoleSessionName")
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
var errResp STSErrorTestResponse
err = xml.Unmarshal(body, &errResp)
require.NoError(t, err, "Failed to parse error response: %s", string(body))
assert.Equal(t, "MissingParameter", errResp.Error.Code)
})
t.Run("unsupported_action_for_anonymous", func(t *testing.T) {
// AssumeRole requires SigV4 authentication, anonymous requests should fail
resp, err := callSTSAPI(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/test-role"},
"RoleSessionName": {"test-session"},
})
require.NoError(t, err)
defer resp.Body.Close()
// Should fail because AssumeRole requires AWS SigV4 authentication
assert.NotEqual(t, http.StatusOK, resp.StatusCode,
"AssumeRole should require authentication")
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
t.Logf("Response for anonymous AssumeRole: status=%d, body=%s", resp.StatusCode, string(body))
})
t.Run("invalid_duration_too_short", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/test-role"},
"RoleSessionName": {"test-session"},
"DurationSeconds": {"100"}, // Less than 900 seconds minimum
}, "test-access-key", "test-secret-key")
require.NoError(t, err)
defer resp.Body.Close()
assert.NotEqual(t, http.StatusOK, resp.StatusCode,
"Should fail with DurationSeconds < 900")
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
var errResp STSErrorTestResponse
err = xml.Unmarshal(body, &errResp)
require.NoError(t, err, "Failed to parse error response: %s", string(body))
assert.Equal(t, "InvalidParameterValue", errResp.Error.Code)
})
t.Run("invalid_duration_too_long", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/test-role"},
"RoleSessionName": {"test-session"},
"DurationSeconds": {"100000"}, // More than 43200 seconds maximum
}, "test-access-key", "test-secret-key")
require.NoError(t, err)
defer resp.Body.Close()
assert.NotEqual(t, http.StatusOK, resp.StatusCode,
"Should fail with DurationSeconds > 43200")
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
var errResp STSErrorTestResponse
err = xml.Unmarshal(body, &errResp)
require.NoError(t, err, "Failed to parse error response: %s", string(body))
assert.Equal(t, "InvalidParameterValue", errResp.Error.Code)
})
}
// isAssumeRoleImplemented checks if the running server supports AssumeRole
func isAssumeRoleImplemented(t *testing.T) bool {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/test"},
"RoleSessionName": {"test"},
}, "test", "test")
if err != nil {
return false
}
defer resp.Body.Close()
body, err := io.ReadAll(resp.Body)
if err != nil {
return false
}
// If we get "NotImplemented", the action isn't supported
var errResp STSErrorTestResponse
if xml.Unmarshal(body, &errResp) == nil && errResp.Error.Code == "NotImplemented" {
return false
}
// If we get InvalidAction, the action isn't routed
if errResp.Error.Code == "InvalidAction" {
return false
}
return true
}
// TestSTSAssumeRoleWithValidCredentials tests AssumeRole with valid IAM credentials
// This test requires a configured IAM user in SeaweedFS
func TestSTSAssumeRoleWithValidCredentials(t *testing.T) {
if testing.Short() {
t.Skip("Skipping integration test in short mode")
}
if !isSTSEndpointRunning(t) {
t.Skip("SeaweedFS STS endpoint is not running at", TestSTSEndpoint)
}
// Use test credentials from environment or fall back to defaults
accessKey := os.Getenv("STS_TEST_ACCESS_KEY")
if accessKey == "" {
accessKey = "admin"
}
secretKey := os.Getenv("STS_TEST_SECRET_KEY")
if secretKey == "" {
secretKey = "admin"
}
t.Run("successful_assume_role", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/admin"},
"RoleSessionName": {"integration-test-session"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
t.Logf("Response status: %d, body: %s", resp.StatusCode, string(body))
// If AssumeRole is not yet implemented, expect an error about unsupported action
if resp.StatusCode != http.StatusOK {
var errResp STSErrorTestResponse
err = xml.Unmarshal(body, &errResp)
require.NoError(t, err, "Failed to parse error response: %s", string(body))
t.Logf("Error response: code=%s, message=%s", errResp.Error.Code, errResp.Error.Message)
// This test will initially fail until AssumeRole is implemented
// Once implemented, uncomment the assertions below
// assert.Fail(t, "AssumeRole not yet implemented")
} else {
var stsResp AssumeRoleTestResponse
err = xml.Unmarshal(body, &stsResp)
require.NoError(t, err, "Failed to parse response: %s", string(body))
creds := stsResp.Result.Credentials
assert.NotEmpty(t, creds.AccessKeyId, "AccessKeyId should not be empty")
assert.NotEmpty(t, creds.SecretAccessKey, "SecretAccessKey should not be empty")
assert.NotEmpty(t, creds.SessionToken, "SessionToken should not be empty")
assert.NotEmpty(t, creds.Expiration, "Expiration should not be empty")
t.Logf("Successfully obtained temporary credentials: AccessKeyId=%s", creds.AccessKeyId)
}
})
t.Run("with_custom_duration", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/admin"},
"RoleSessionName": {"duration-test-session"},
"DurationSeconds": {"3600"}, // 1 hour
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
t.Logf("Response status: %d, body: %s", resp.StatusCode, string(body))
// Verify DurationSeconds is accepted
if resp.StatusCode != http.StatusOK {
var errResp STSErrorTestResponse
err = xml.Unmarshal(body, &errResp)
require.NoError(t, err, "Failed to parse error response: %s", string(body))
// Should not fail due to DurationSeconds parameter
assert.NotContains(t, errResp.Error.Message, "DurationSeconds",
"DurationSeconds parameter should be accepted")
}
})
}
// TestSTSAssumeRoleWithInvalidCredentials tests AssumeRole rejection with bad credentials
func TestSTSAssumeRoleWithInvalidCredentials(t *testing.T) {
if testing.Short() {
t.Skip("Skipping integration test in short mode")
}
if !isSTSEndpointRunning(t) {
t.Skip("SeaweedFS STS endpoint is not running at", TestSTSEndpoint)
}
t.Run("invalid_access_key", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/admin"},
"RoleSessionName": {"test-session"},
}, "invalid-access-key", "some-secret-key")
require.NoError(t, err)
defer resp.Body.Close()
// Should fail with access denied or signature mismatch
assert.NotEqual(t, http.StatusOK, resp.StatusCode,
"Should fail with invalid access key")
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
t.Logf("Response for invalid credentials: status=%d, body=%s", resp.StatusCode, string(body))
})
t.Run("invalid_secret_key", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/admin"},
"RoleSessionName": {"test-session"},
}, "admin", "wrong-secret-key")
require.NoError(t, err)
defer resp.Body.Close()
// Should fail with signature mismatch
assert.NotEqual(t, http.StatusOK, resp.StatusCode,
"Should fail with invalid secret key")
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
t.Logf("Response for wrong secret: status=%d, body=%s", resp.StatusCode, string(body))
})
}
// callSTSAPIWithSigV4 makes an STS API call with AWS Signature V4 authentication
func callSTSAPIWithSigV4(t *testing.T, params url.Values, accessKey, secretKey string) (*http.Response, error) {
// Prepare request body
body := params.Encode()
// Create request
req, err := http.NewRequest(http.MethodPost, TestSTSEndpoint+"/",
strings.NewReader(body))
if err != nil {
return nil, err
}
req.Header.Set("Content-Type", "application/x-www-form-urlencoded")
// Do NOT set req.Header["Host"] — aws-sdk-go v1 v4.Signer reads the host
// from req.URL.Host/req.Host, and a manual Host header makes the signer
// emit `host;host` in SignedHeaders, producing a different signature than
// the server calculates from canonical headers.
// Sign request with AWS Signature V4 using official SDK
creds := credentials.NewStaticCredentials(accessKey, secretKey, "")
signer := v4.NewSigner(creds)
// Read body for signing
// Note: We need a ReadSeeker for the signer, or we can pass the body string/bytes to ComputeBodyHash if needed,
// but standard Sign method takes an io.ReadSeeker for the body.
bodyReader := strings.NewReader(body)
_, err = signer.Sign(req, bodyReader, "sts", "us-east-1", time.Now())
if err != nil {
return nil, fmt.Errorf("failed to sign request: %w", err)
}
client := &http.Client{Timeout: 30 * time.Second}
return client.Do(req)
}