Files
3ff4fbc7e4 fix(eventstore): add owners to unique constraints (#12740)
<!-- CURSOR_AGENT_PR_BODY_BEGIN -->
# Which Problems Are Solved

Bulk unique-constraint deletes (org, IDP with many links, user, project)
reconstruct every uniqueness key in Go and emit N `DELETE`s. That can
OOM on large tenants.

A one-shot owner backfill is also not enough: during a rolling deploy,
older binaries still insert `owners = '{}'`, so a later owner-only
delete would miss those rows.

Owner-only org/user/grant delete only matches rows whose `owners` bag
contains that tag. Several grant and member types stamped the aggregate
resource owner, which is not always the user’s home org or the granted
org. Username-scope rewrites also dropped `user:` tags, so owner-only
user delete could miss rewritten username rows.

Command must not switch to owner-only delete on the first 79 of a
version, because leftover `'{}'` rows from rolling-update overlap are
only backfilled on a later 79. A failed 79 must not turn owner-delete
on, and backfill joins must not stamp the wrong org (unverified domains,
org-scoped usernames colliding with instance-scoped ones).

# How the Problems Are Solved

Uniqueness stays the existing primary key `(instance_id, unique_type,
unique_field)`. `owners` is only a lifecycle bag of tags (`org:<id>`,
`user:<id>`, `idp:<id>`, `project:<id>`, `grant:<id>`).

- Add `owners TEXT[] NOT NULL DEFAULT '{}'` on
`eventstore.unique_constraints` (init SQL for greenfield, setup 78 for
existing clusters).
- Create `unique_constraints_owners_gin` with `CREATE INDEX
CONCURRENTLY` **outside a transaction** (same pattern as step 76).
Built-in `GIN (owners)` only; no `btree_gin`.
- Stamp `WithOwners` on **add** events. Field-level removes ignore
owners. `RemoveByOwner` stores the tag in `Owners` and both v1/v3 and v2
pushers run `DELETE … WHERE instance_id = $1 AND owners @>
ARRAY[$tag]::text[]`.
- Setup 79 backfills `owners` from projections via numbered
`cmd/setup/79/*.sql` files. Table names are interpolated from
`internal/query/projection` constants. Matches use `unique_field IN (x,
lower(x))` so the primary key can be used. Missing projection tables
(`42P01`) are skipped.
- Org domains are tagged only when `is_verified`. Instance-scoped
usernames join only users that are not org-scoped (organization settings
or domain policy `user_login_must_be_domain`). Every username row still
gets `org:` + `user:` of that user.
- 79 is a **repeatable** migration. `Check` re-runs when
`lastRun["version"]` changed or when setup-only `ForceFinalize` is set
and `lastRun["finalized"]` is still false. `Finalized` is written only
after Execute succeeds.
- Failed setup events keep the last successful `lastRun`. Command reads
`finalized` only from `system.migration.repeatable.done`.
- Org / user / project remove skip N-key event replay **when that
`finalized` bit is true**. Otherwise they keep today’s N-key lists
**and** still emit `RemoveByOwner`. The “ready” result is cached in
memory once true.

Setup writes `finalized` on the 79 repeatable.done payload:

- First 79 of a version: backfill only, `finalized=false` (unless
`ForceFinalize` is already set).
- Later 79: version change or
`BackfillUniqueConstraintOwners.ForceFinalize` /
`ZITADEL_BACKFILLUNIQUECONSTRAINTOWNERS_FORCEFINALIZE` after the rolling
update is done. That run backfills leftovers and writes
`finalized=true`.

Member and grant owner bags now include every org that should delete the
row:

- Instance, org, and project members also stamp
`org:<user_resource_owner>`.
- User grants stamp grant, user, project, and granted orgs.
- Project grant members stamp the project org, user org, and granted
org.
- Project grants stamp both orgs plus `grant:<id>`. Grant remove emits
`RemoveByOwner(grant)`.
- Username-scope rewrites keep `user:<id>` as well as `org:`.
- User/grant orgs are carried as unexported add-event fields (not
persisted JSON). `WithOwners` skips empty and duplicate tags.

# Additional Changes

- Table-driven owner-tag tests, including project grant (two `org:` tags
plus `grant:`), cross-org members, user grants, and username-scope
rewrite `user:` tags.
- Table-driven 79 `Check` + `finalizeAfterSuccess` tests for version
change, sequential version bumps, ForceFinalize restart, and
already-finalized skip.
- Command tests: a failed 79 event with `finalized=true` does not enable
owner-delete.
- Integration fixtures (verified vs unverified org domain;
instance-scoped `bob` vs org-scoped `bob`).
- Instance-only and global unique types stay untagged (`'{}'`). Instance
wipe still uses `UniqueConstraintInstanceRemove`.
- Pusher owner cases live in `eventstore_pusher_owners_test.go` so
`eventstore_pusher_test.go` stays under 1k lines.

# Additional Context

A later cleanup can delete the N-key replay helpers once a release that
ran 79 a second time (or an operator who set `ForceFinalize`) is in
production. That is not required for correctness.

<!-- CURSOR_AGENT_PR_BODY_END -->

<div><a
href="https://cursor.com/agents/bc-44f9bc9b-9a93-4cd9-8dd1-663a7b9aad2a?cursor_ref=pr_footer&cursor_cta=open_in_web"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://cursor.com/assets/images/open-in-web-dark.png"><source
media="(prefers-color-scheme: light)"
srcset="https://cursor.com/assets/images/open-in-web-light.png"><img
alt="Open in Web" width="114" height="28"
src="https://cursor.com/assets/images/open-in-web-dark.png"></picture></a>&nbsp;<a
href="https://cursor.com/background-agent?bcId=bc-44f9bc9b-9a93-4cd9-8dd1-663a7b9aad2a&cursor_ref=pr_footer&cursor_cta=open_in_cursor"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://cursor.com/assets/images/open-in-cursor-dark.png"><source
media="(prefers-color-scheme: light)"
srcset="https://cursor.com/assets/images/open-in-cursor-light.png"><img
alt="Open in Cursor" width="131" height="28"
src="https://cursor.com/assets/images/open-in-cursor-dark.png"></picture></a>&nbsp;</div>

---------

Co-authored-by: Cursor Agent <[email protected]>
Co-authored-by: Silvan <[email protected]>
2026-09-18 15:08:42 +02:00

408 lines
15 KiB
Go

package command
import (
"context"
"crypto/rand"
"crypto/x509"
"crypto/x509/pkix"
"encoding/pem"
"fmt"
"math/big"
"net"
"net/http"
"slices"
"strconv"
"strings"
"sync"
"sync/atomic"
"time"
"github.com/go-jose/go-jose/v4"
"github.com/zitadel/logging"
new_domain "github.com/zitadel/zitadel/backend/v3/domain"
"github.com/zitadel/zitadel/internal/api/authz"
api_http "github.com/zitadel/zitadel/internal/api/http"
"github.com/zitadel/zitadel/internal/cache/connector"
"github.com/zitadel/zitadel/internal/command/preparation"
sd "github.com/zitadel/zitadel/internal/config/systemdefaults"
"github.com/zitadel/zitadel/internal/crypto"
"github.com/zitadel/zitadel/internal/denylist"
"github.com/zitadel/zitadel/internal/domain"
"github.com/zitadel/zitadel/internal/eventstore"
"github.com/zitadel/zitadel/internal/id"
internal_net "github.com/zitadel/zitadel/internal/net"
"github.com/zitadel/zitadel/internal/notification/senders"
"github.com/zitadel/zitadel/internal/static"
"github.com/zitadel/zitadel/internal/telemetry/tracing"
webauthn_helper "github.com/zitadel/zitadel/internal/webauthn"
"github.com/zitadel/zitadel/internal/zerrors"
)
type Commands struct {
httpClient *http.Client
jobs sync.WaitGroup
checkPermission domain.PermissionCheck
newEncryptedCode encrypedCodeFunc
newEncryptedCodeWithDefault encryptedCodeWithDefaultFunc
newHashedSecret hashedSecretFunc
eventstore *eventstore.Eventstore
static static.Storage
idGenerator id.Generator
zitadelRoles []authz.RoleMapping
externalDomain string
externalSecure bool
externalPort uint16
idpConfigEncryption crypto.AuthEncryptionAlgorithm
smtpEncryption crypto.EncryptionAlgorithm
smsEncryption crypto.EncryptionAlgorithm
userEncryption crypto.EncryptionAlgorithm
targetEncryption crypto.EncryptionAlgorithm
userPasswordHasher *crypto.Hasher
secretHasher *crypto.Hasher
machineKeySize int
applicationKeySize int
domainVerificationAlg crypto.EncryptionAlgorithm
domainVerificationGenerator crypto.Generator
domainVerificationValidator func(domain, token, verifier string, checkType api_http.CheckType) error
sessionTokenCreator func(sessionID string) (id string, token string, err error)
sessionTokenVerifier func(ctx context.Context, sessionToken, sessionID, tokenID string) (err error)
defaultAccessTokenLifetime time.Duration
defaultRefreshTokenLifetime time.Duration
defaultRefreshTokenIdleLifetime time.Duration
phoneCodeVerifier func(ctx context.Context, id string) (senders.CodeGenerator, error)
tarpit func(failedAttempts uint64)
multifactors domain.MultifactorConfigs
webauthnConfig *webauthn_helper.Config
keySize int
keyAlgorithm crypto.EncryptionAlgorithm
authAlgorithm crypto.AuthAlgorithm
certificateAlgorithm crypto.EncryptionAlgorithm
certKeySize int
privateKeyLifetime time.Duration
publicKeyLifetime time.Duration
certificateLifetime time.Duration
defaultSecretGenerators *SecretGenerators
maxIdPIntentLifetime time.Duration
samlCertificateAndKeyGenerator func(id string) ([]byte, []byte, error)
webKeyGenerator func(keyID string, alg crypto.EncryptionAlgorithm, genConfig crypto.WebKeyConfig) (encryptedPrivate *crypto.CryptoValue, public *jose.JSONWebKey, err error)
GrpcMethodExisting func(method string) bool
GrpcServiceExisting func(method string) bool
ActionFunctionExisting func(function string) bool
EventExisting func(event string) bool
EventGroupExisting func(group string) bool
GenerateDomain func(instanceName, domain string) (string, error)
caches *Caches
// Store instance IDs where all milestones are reached (except InstanceDeleted).
// These instance's milestones never need to be invalidated,
// so the query and cache overhead can be completely eliminated.
milestonesCompleted sync.Map
loginPaths LoginPaths
ipLookupFunction internal_net.IPLookupFunc
denyList []denylist.AddressChecker
ownerDeleteReady func(context.Context) (bool, error)
ownerDeleteReadyCached atomic.Bool
}
//go:generate mockgen -package command -destination ./mock_login_paths.go . LoginPaths
type LoginPaths interface {
DefaultEmailCodeURLTemplate(ctx context.Context) string
DefaultPasswordSetURLTemplate(ctx context.Context) string
DefaultPasskeySetURLTemplate(ctx context.Context) string
DefaultDomainClaimedURLTemplate(ctx context.Context) string
}
func StartCommands(
ctx context.Context,
es *eventstore.Eventstore,
cacheConnectors connector.Connectors,
defaults sd.SystemDefaults,
zitadelRoles []authz.RoleMapping,
staticStore static.Storage,
webAuthN *webauthn_helper.Config,
externalDomain string,
externalSecure bool,
externalPort uint16,
idpConfigEncryption crypto.AuthEncryptionAlgorithm,
otpEncryption, smtpEncryption, smsEncryption, userEncryption, domainVerificationEncryption, samlEncryption, targetEncryption crypto.EncryptionAlgorithm,
oidcEncryption crypto.AuthEncryptionAlgorithm,
httpClient *http.Client,
permissionCheck domain.PermissionCheck,
sessionTokenVerifier func(ctx context.Context, sessionToken string, sessionID string, tokenID string) (err error),
defaultAccessTokenLifetime, defaultRefreshTokenLifetime, defaultRefreshTokenIdleLifetime time.Duration,
defaultSecretGenerators *SecretGenerators,
loginPaths LoginPaths,
denyList []denylist.AddressChecker,
) (repo *Commands, err error) {
if externalDomain == "" {
return nil, zerrors.ThrowInvalidArgument(nil, "COMMAND-Df21s", "no external domain specified")
}
idGenerator := id.SonyFlakeGenerator()
// reuse the oidcEncryption to be able to handle both tokens in the interceptor later on
sessionAlg := oidcEncryption
secretHasher, err := defaults.SecretHasher.NewHasher()
if err != nil {
return nil, fmt.Errorf("secret hasher: %w", err)
}
userPasswordHasher, err := defaults.PasswordHasher.NewHasher()
if err != nil {
return nil, fmt.Errorf("password hasher: %w", err)
}
new_domain.SetPasswordHasher(userPasswordHasher)
caches, err := startCaches(ctx, cacheConnectors)
if err != nil {
return nil, fmt.Errorf("caches: %w", err)
}
ipLookupFunction := net.LookupIP
repo = &Commands{
eventstore: es,
static: staticStore,
idGenerator: idGenerator,
zitadelRoles: zitadelRoles,
externalDomain: externalDomain,
externalSecure: externalSecure,
externalPort: externalPort,
keySize: defaults.KeyConfig.Size,
certKeySize: defaults.KeyConfig.CertificateSize,
privateKeyLifetime: defaults.KeyConfig.PrivateKeyLifetime,
publicKeyLifetime: defaults.KeyConfig.PublicKeyLifetime,
certificateLifetime: defaults.KeyConfig.CertificateLifetime,
maxIdPIntentLifetime: defaults.MaxIdPIntentLifetime,
idpConfigEncryption: idpConfigEncryption,
smtpEncryption: smtpEncryption,
smsEncryption: smsEncryption,
userEncryption: userEncryption,
targetEncryption: targetEncryption,
userPasswordHasher: userPasswordHasher,
secretHasher: secretHasher,
machineKeySize: int(defaults.SecretGenerators.MachineKeySize),
applicationKeySize: int(defaults.SecretGenerators.ApplicationKeySize),
domainVerificationAlg: domainVerificationEncryption,
domainVerificationGenerator: crypto.NewEncryptionGenerator(defaults.DomainVerification.VerificationGenerator, domainVerificationEncryption),
domainVerificationValidator: func(domain, token, verifier string, checkType api_http.CheckType) error {
return api_http.ValidateDomain(domain, token, verifier, checkType, httpClient)
},
keyAlgorithm: oidcEncryption,
authAlgorithm: oidcEncryption,
certificateAlgorithm: samlEncryption,
webauthnConfig: webAuthN,
httpClient: httpClient,
checkPermission: permissionCheck,
newEncryptedCode: newEncryptedCode,
newEncryptedCodeWithDefault: newEncryptedCodeWithDefaultConfig,
sessionTokenCreator: sessionTokenCreator(idGenerator, sessionAlg),
sessionTokenVerifier: sessionTokenVerifier,
defaultAccessTokenLifetime: defaultAccessTokenLifetime,
defaultRefreshTokenLifetime: defaultRefreshTokenLifetime,
defaultRefreshTokenIdleLifetime: defaultRefreshTokenIdleLifetime,
defaultSecretGenerators: defaultSecretGenerators,
samlCertificateAndKeyGenerator: samlCertificateAndKeyGenerator(defaults.KeyConfig.CertificateSize, defaults.KeyConfig.CertificateLifetime),
webKeyGenerator: crypto.GenerateEncryptedWebKey,
EventExisting: func(value string) bool {
return slices.Contains(es.EventTypes(), value)
},
EventGroupExisting: func(group string) bool {
return slices.ContainsFunc(es.EventTypes(), func(value string) bool {
return strings.HasPrefix(value, group)
},
)
},
GrpcServiceExisting: func(service string) bool { return false },
GrpcMethodExisting: func(method string) bool { return false },
ActionFunctionExisting: domain.ActionFunctionExists(),
multifactors: domain.MultifactorConfigs{
OTP: domain.OTPConfig{
CryptoMFA: otpEncryption,
Issuer: defaults.Multifactors.OTP.Issuer,
},
RecoveryCodes: domain.RecoveryCodesConfig{
MaxCount: defaults.Multifactors.RecoveryCodes.MaxCount,
Format: domain.RecoveryCodeFormat(defaults.Multifactors.RecoveryCodes.Format),
Length: defaults.Multifactors.RecoveryCodes.Length,
WithHyphen: defaults.Multifactors.RecoveryCodes.WithHyphen,
},
},
GenerateDomain: domain.NewGeneratedInstanceDomain,
caches: caches,
loginPaths: loginPaths,
ipLookupFunction: ipLookupFunction,
denyList: denyList,
}
repo.ownerDeleteReady = repo.uniqueConstraintOwnersBackfillFinalized
if defaultSecretGenerators != nil && defaultSecretGenerators.ClientSecret != nil {
repo.newHashedSecret = newHashedSecretWithDefault(secretHasher, defaultSecretGenerators.ClientSecret)
}
repo.phoneCodeVerifier = repo.phoneCodeVerifierFromConfig
repo.tarpit = defaults.Tarpit.Tarpit()
return repo, nil
}
type AppendReducer interface {
AppendEvents(...eventstore.Event)
// TODO: Why is it allowed to return an error here?
Reduce() error
}
func (c *Commands) pushAppendAndReduce(ctx context.Context, object AppendReducer, cmds ...eventstore.Command) error {
if len(cmds) == 0 {
return nil
}
events, err := c.eventstore.Push(ctx, cmds...)
if err != nil {
return err
}
return AppendAndReduce(object, events...)
}
type AppendReducerDetails interface {
AppendEvents(...eventstore.Event)
// TODO: Why is it allowed to return an error here?
Reduce() error
GetWriteModel() *eventstore.WriteModel
}
func (c *Commands) pushAppendAndReduceDetails(ctx context.Context, object AppendReducerDetails, cmds ...eventstore.Command) (*domain.ObjectDetails, error) {
if err := c.pushAppendAndReduce(ctx, object, cmds...); err != nil {
return nil, err
}
return writeModelToObjectDetails(object.GetWriteModel()), nil
}
func AppendAndReduce(object AppendReducer, events ...eventstore.Event) error {
object.AppendEvents(events...)
return object.Reduce()
}
func queryAndReduce(ctx context.Context, filter preparation.FilterToQueryReducer, wm eventstore.QueryReducer) error {
events, err := filter(ctx, wm.Query())
if err != nil {
return err
}
if len(events) == 0 {
return nil
}
wm.AppendEvents(events...)
return wm.Reduce()
}
type existsWriteModel interface {
Exists() bool
eventstore.QueryReducer
}
func exists(ctx context.Context, filter preparation.FilterToQueryReducer, wm existsWriteModel) (bool, error) {
err := queryAndReduce(ctx, filter, wm)
if err != nil {
return false, err
}
return wm.Exists(), nil
}
func samlCertificateAndKeyGenerator(keySize int, lifetime time.Duration) func(id string) ([]byte, []byte, error) {
return func(id string) ([]byte, []byte, error) {
priv, pub, err := crypto.GenerateKeyPair(keySize)
if err != nil {
return nil, nil, err
}
serial, err := strconv.Atoi(id)
if err != nil {
return nil, nil, err
}
now := time.Now()
template := x509.Certificate{
SerialNumber: big.NewInt(int64(serial)),
Subject: pkix.Name{
Organization: []string{"ZITADEL"},
CommonName: fmt.Sprintf("ZITADEL SP %s", id),
SerialNumber: id,
},
Issuer: pkix.Name{
Organization: []string{"ZITADEL"},
CommonName: "ZITADEL",
},
NotBefore: now,
NotAfter: now.Add(lifetime),
KeyUsage: x509.KeyUsageKeyEncipherment | x509.KeyUsageDigitalSignature,
ExtKeyUsage: []x509.ExtKeyUsage{x509.ExtKeyUsageServerAuth},
BasicConstraintsValid: true,
}
derBytes, err := x509.CreateCertificate(rand.Reader, &template, &template, pub, priv)
if err != nil {
return nil, nil, zerrors.ThrowInternalf(err, "COMMAND-x92u101j", "failed to create certificate")
}
keyBlock := &pem.Block{Type: "RSA PRIVATE KEY", Bytes: x509.MarshalPKCS1PrivateKey(priv)}
certBlock := &pem.Block{Type: "CERTIFICATE", Bytes: derBytes}
return pem.EncodeToMemory(keyBlock), pem.EncodeToMemory(certBlock), nil
}
}
// Close blocks until all async jobs are finished,
// the context expires or after eventstore.PushTimeout.
func (c *Commands) Close(ctx context.Context) error {
if c.eventstore.PushTimeout != 0 {
var cancel context.CancelFunc
ctx, cancel = context.WithTimeout(ctx, c.eventstore.PushTimeout)
defer cancel()
}
done := make(chan struct{})
go func() {
c.jobs.Wait()
close(done)
}()
select {
case <-done:
return nil
case <-ctx.Done():
return ctx.Err()
}
}
// asyncPush attempts to push events to the eventstore in a separate Go routine.
// This can be used to speed up request times when the outcome of the push is
// not important for business logic but have a pure logging function.
// For example this can be used for Secret Check Success and Failed events.
// On push error, a log line describing the error will be emitted.
func (c *Commands) asyncPush(ctx context.Context, cmds ...eventstore.Command) {
// Create a new context, as the request scoped context might get
// canceled before we where able to push.
// The eventstore has its own PushTimeout setting,
// so we don't need to have a context with timeout here.
ctx = context.WithoutCancel(ctx)
c.jobs.Add(1)
go func() {
defer c.jobs.Done()
localCtx, span := tracing.NewSpan(ctx)
_, err := c.eventstore.Push(localCtx, cmds...)
if err != nil {
for _, cmd := range cmds {
logging.WithError(err).Warnf("could not push event %q", cmd.Type())
}
}
span.EndWithError(err)
}()
}