You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
consul/agent/leafcert/signer_test.go

244 lines
6.1 KiB

agent: remove agent cache dependency from service mesh leaf certificate management (#17075) * agent: remove agent cache dependency from service mesh leaf certificate management This extracts the leaf cert management from within the agent cache. This code was produced by the following process: 1. All tests in agent/cache, agent/cache-types, agent/auto-config, agent/consul/servercert were run at each stage. - The tests in agent matching .*Leaf were run at each stage. - The tests in agent/leafcert were run at each stage after they existed. 2. The former leaf cert Fetch implementation was extracted into a new package behind a "fake RPC" endpoint to make it look almost like all other cache type internals. 3. The old cache type was shimmed to use the fake RPC endpoint and generally cleaned up. 4. I selectively duplicated all of Get/Notify/NotifyCallback/Prepopulate from the agent/cache.Cache implementation over into the new package. This was renamed as leafcert.Manager. - Code that was irrelevant to the leaf cert type was deleted (inlining blocking=true, refresh=false) 5. Everything that used the leaf cert cache type (including proxycfg stuff) was shifted to use the leafcert.Manager instead. 6. agent/cache-types tests were moved and gently replumbed to execute as-is against a leafcert.Manager. 7. Inspired by some of the locking changes from derek's branch I split the fat lock into N+1 locks. 8. The waiter chan struct{} was eventually replaced with a singleflight.Group around cache updates, which was likely the biggest net structural change. 9. The awkward two layers or logic produced as a byproduct of marrying the agent cache management code with the leaf cert type code was slowly coalesced and flattened to remove confusion. 10. The .*Leaf tests from the agent package were copied and made to work directly against a leafcert.Manager to increase direct coverage. I have done a best effort attempt to port the previous leaf-cert cache type's tests over in spirit, as well as to take the e2e-ish tests in the agent package with Leaf in the test name and copy those into the agent/leafcert package to get more direct coverage, rather than coverage tangled up in the agent logic. There is no net-new test coverage, just coverage that was pushed around from elsewhere.
1 year ago
package leafcert
import (
"bytes"
"context"
"crypto/rand"
"crypto/x509"
"encoding/pem"
"errors"
"fmt"
"math/big"
"sync"
"sync/atomic"
"testing"
"time"
"github.com/hashicorp/consul/agent/connect"
"github.com/hashicorp/consul/agent/structs"
)
// testSigner implements NetRPC and handles leaf signing operations
type testSigner struct {
caLock sync.Mutex
ca *structs.CARoot
prevRoots []*structs.CARoot // remember prior ones
IDGenerator *atomic.Uint64
RootsReader *testRootsReader
signCallLock sync.Mutex
signCallErrors []error
signCallErrorCount uint64
signCallCapture []*structs.CASignRequest
}
var _ CertSigner = (*testSigner)(nil)
var ReplyWithExpiredCert = errors.New("reply with expired cert")
func newTestSigner(t *testing.T, idGenerator *atomic.Uint64, rootsReader *testRootsReader) *testSigner {
if idGenerator == nil {
idGenerator = &atomic.Uint64{}
}
if rootsReader == nil {
rootsReader = newTestRootsReader(t)
}
s := &testSigner{
IDGenerator: idGenerator,
RootsReader: rootsReader,
}
return s
}
func (s *testSigner) SetSignCallErrors(errs ...error) {
s.signCallLock.Lock()
defer s.signCallLock.Unlock()
s.signCallErrors = append(s.signCallErrors, errs...)
}
func (s *testSigner) GetSignCallErrorCount() uint64 {
s.signCallLock.Lock()
defer s.signCallLock.Unlock()
return s.signCallErrorCount
}
func (s *testSigner) UpdateCA(t *testing.T, ca *structs.CARoot) *structs.CARoot {
if ca == nil {
ca = connect.TestCA(t, nil)
}
roots := &structs.IndexedCARoots{
ActiveRootID: ca.ID,
TrustDomain: connect.TestTrustDomain,
Roots: []*structs.CARoot{ca},
QueryMeta: structs.QueryMeta{Index: s.nextIndex()},
}
// Update the signer first.
s.caLock.Lock()
{
s.ca = ca
roots.Roots = append(roots.Roots, s.prevRoots...)
// Remember for the next rotation.
dup := ca.Clone()
dup.Active = false
s.prevRoots = append(s.prevRoots, dup)
}
s.caLock.Unlock()
// Then trigger an event when updating the roots.
s.RootsReader.Set(roots)
return ca
}
func (s *testSigner) nextIndex() uint64 {
return s.IDGenerator.Add(1)
}
func (s *testSigner) getCA() *structs.CARoot {
s.caLock.Lock()
defer s.caLock.Unlock()
return s.ca
}
func (s *testSigner) GetCapture(idx int) *structs.CASignRequest {
s.signCallLock.Lock()
defer s.signCallLock.Unlock()
if len(s.signCallCapture) > idx {
return s.signCallCapture[idx]
}
return nil
}
func (s *testSigner) SignCert(ctx context.Context, req *structs.CASignRequest) (*structs.IssuedCert, error) {
useExpiredCert := false
s.signCallLock.Lock()
s.signCallCapture = append(s.signCallCapture, req)
if len(s.signCallErrors) > 0 {
err := s.signCallErrors[0]
s.signCallErrors = s.signCallErrors[1:]
if err == ReplyWithExpiredCert {
useExpiredCert = true
} else if err != nil {
s.signCallErrorCount++
s.signCallLock.Unlock()
return nil, err
}
}
s.signCallLock.Unlock()
// parts of this were inlined from CAManager and the connect ca provider
ca := s.getCA()
if ca == nil {
return nil, fmt.Errorf("must call UpdateCA at least once")
}
csr, err := connect.ParseCSR(req.CSR)
if err != nil {
return nil, fmt.Errorf("error parsing CSR: %w", err)
}
connect.HackSANExtensionForCSR(csr)
spiffeID, err := connect.ParseCertURI(csr.URIs[0])
if err != nil {
return nil, fmt.Errorf("error parsing CSR URI: %w", err)
}
serviceID, isService := spiffeID.(*connect.SpiffeIDService)
if !isService {
return nil, fmt.Errorf("unexpected spiffeID type %T", spiffeID)
}
signer, err := connect.ParseSigner(ca.SigningKey)
if err != nil {
return nil, fmt.Errorf("error parsing CA signing key: %w", err)
}
keyId, err := connect.KeyId(signer.Public())
if err != nil {
return nil, fmt.Errorf("error forming CA key id from public key: %w", err)
}
subjectKeyID, err := connect.KeyId(csr.PublicKey)
if err != nil {
return nil, fmt.Errorf("error forming subject key id from public key: %w", err)
}
caCert, err := connect.ParseCert(ca.RootCert)
if err != nil {
return nil, fmt.Errorf("error parsing CA root cert pem: %w", err)
}
const expiration = 10 * time.Minute
now := time.Now()
template := x509.Certificate{
SerialNumber: big.NewInt(int64(s.nextIndex())),
URIs: csr.URIs,
Signature: csr.Signature,
// We use the correct signature algorithm for the CA key we are signing with
// regardless of the algorithm used to sign the CSR signature above since
// the leaf might use a different key type.
SignatureAlgorithm: connect.SigAlgoForKey(signer),
PublicKeyAlgorithm: csr.PublicKeyAlgorithm,
PublicKey: csr.PublicKey,
BasicConstraintsValid: true,
KeyUsage: x509.KeyUsageDataEncipherment |
x509.KeyUsageKeyAgreement |
x509.KeyUsageDigitalSignature |
x509.KeyUsageKeyEncipherment,
ExtKeyUsage: []x509.ExtKeyUsage{
x509.ExtKeyUsageClientAuth,
x509.ExtKeyUsageServerAuth,
},
NotAfter: now.Add(expiration),
NotBefore: now,
AuthorityKeyId: keyId,
SubjectKeyId: subjectKeyID,
DNSNames: csr.DNSNames,
IPAddresses: csr.IPAddresses,
}
if useExpiredCert {
template.NotBefore = time.Now().Add(-13 * time.Hour)
template.NotAfter = time.Now().Add(-1 * time.Hour)
}
// Create the certificate, PEM encode it and return that value.
var buf bytes.Buffer
bs, err := x509.CreateCertificate(
rand.Reader, &template, caCert, csr.PublicKey, signer)
if err != nil {
return nil, fmt.Errorf("error creating cert pem from CSR: %w", err)
}
err = pem.Encode(&buf, &pem.Block{Type: "CERTIFICATE", Bytes: bs})
if err != nil {
return nil, fmt.Errorf("error encoding cert pem into text: %w", err)
}
leafPEM := buf.String()
leafCert, err := connect.ParseCert(leafPEM)
if err != nil {
return nil, fmt.Errorf("error parsing cert from generated leaf pem: %w", err)
}
index := s.nextIndex()
return &structs.IssuedCert{
SerialNumber: connect.EncodeSerialNumber(leafCert.SerialNumber),
CertPEM: leafPEM,
Service: serviceID.Service,
ServiceURI: leafCert.URIs[0].String(),
ValidAfter: leafCert.NotBefore,
ValidBefore: leafCert.NotAfter,
RaftIndex: structs.RaftIndex{
CreateIndex: index,
ModifyIndex: index,
},
}, nil
}