Files
zitadel/internal/eventstore/v3/eventstore.go

218 lines
4.8 KiB
Go
Raw Normal View History

package eventstore
import (
"context"
refactor(eventstore): move push logic to sql (#8816) # Which Problems Are Solved If many events are written to the same aggregate id it can happen that zitadel [starts to retry the push transaction](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/eventstore.go#L101) because [the locking behaviour](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/v3/sequence.go#L25) during push does compute the wrong sequence because newly committed events are not visible to the transaction. These events impact the current sequence. In cases with high command traffic on a single aggregate id this can have severe impact on general performance of zitadel. Because many connections of the `eventstore pusher` database pool are blocked by each other. # How the Problems Are Solved To improve the performance this locking mechanism was removed and the business logic of push is moved to sql functions which reduce network traffic and can be analyzed by the database before the actual push. For clients of the eventstore framework nothing changed. # Additional Changes - after a connection is established prefetches the newly added database types - `eventstore.BaseEvent` now returns the correct revision of the event # Additional Context - part of https://github.com/zitadel/zitadel/issues/8931 --------- Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Miguel Cabrerizo <30386061+doncicuto@users.noreply.github.com> Co-authored-by: Joakim Lodén <Loddan@users.noreply.github.com> Co-authored-by: Yxnt <Yxnt@users.noreply.github.com> Co-authored-by: Stefan Benz <stefan@caos.ch> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Zach H <zhirschtritt@gmail.com>
2024-12-04 14:51:40 +01:00
"database/sql"
"encoding/json"
"errors"
"sync"
"github.com/DATA-DOG/go-sqlmock"
"github.com/jackc/pgx/v5"
"github.com/jackc/pgx/v5/pgtype"
"github.com/jackc/pgx/v5/stdlib"
"github.com/zitadel/logging"
"github.com/zitadel/zitadel/internal/database"
refactor(eventstore): move push logic to sql (#8816) # Which Problems Are Solved If many events are written to the same aggregate id it can happen that zitadel [starts to retry the push transaction](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/eventstore.go#L101) because [the locking behaviour](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/v3/sequence.go#L25) during push does compute the wrong sequence because newly committed events are not visible to the transaction. These events impact the current sequence. In cases with high command traffic on a single aggregate id this can have severe impact on general performance of zitadel. Because many connections of the `eventstore pusher` database pool are blocked by each other. # How the Problems Are Solved To improve the performance this locking mechanism was removed and the business logic of push is moved to sql functions which reduce network traffic and can be analyzed by the database before the actual push. For clients of the eventstore framework nothing changed. # Additional Changes - after a connection is established prefetches the newly added database types - `eventstore.BaseEvent` now returns the correct revision of the event # Additional Context - part of https://github.com/zitadel/zitadel/issues/8931 --------- Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Miguel Cabrerizo <30386061+doncicuto@users.noreply.github.com> Co-authored-by: Joakim Lodén <Loddan@users.noreply.github.com> Co-authored-by: Yxnt <Yxnt@users.noreply.github.com> Co-authored-by: Stefan Benz <stefan@caos.ch> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Zach H <zhirschtritt@gmail.com>
2024-12-04 14:51:40 +01:00
"github.com/zitadel/zitadel/internal/database/dialect"
"github.com/zitadel/zitadel/internal/eventstore"
)
refactor(eventstore): move push logic to sql (#8816) # Which Problems Are Solved If many events are written to the same aggregate id it can happen that zitadel [starts to retry the push transaction](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/eventstore.go#L101) because [the locking behaviour](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/v3/sequence.go#L25) during push does compute the wrong sequence because newly committed events are not visible to the transaction. These events impact the current sequence. In cases with high command traffic on a single aggregate id this can have severe impact on general performance of zitadel. Because many connections of the `eventstore pusher` database pool are blocked by each other. # How the Problems Are Solved To improve the performance this locking mechanism was removed and the business logic of push is moved to sql functions which reduce network traffic and can be analyzed by the database before the actual push. For clients of the eventstore framework nothing changed. # Additional Changes - after a connection is established prefetches the newly added database types - `eventstore.BaseEvent` now returns the correct revision of the event # Additional Context - part of https://github.com/zitadel/zitadel/issues/8931 --------- Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Miguel Cabrerizo <30386061+doncicuto@users.noreply.github.com> Co-authored-by: Joakim Lodén <Loddan@users.noreply.github.com> Co-authored-by: Yxnt <Yxnt@users.noreply.github.com> Co-authored-by: Stefan Benz <stefan@caos.ch> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Zach H <zhirschtritt@gmail.com>
2024-12-04 14:51:40 +01:00
func init() {
dialect.RegisterAfterConnect(RegisterEventstoreTypes)
}
var (
// pushPlaceholderFmt defines how data are inserted into the events table
chore!: Introduce ZITADEL v3 (#9645) This PR summarizes multiple changes specifically only available with ZITADEL v3: - feat: Web Keys management (https://github.com/zitadel/zitadel/pull/9526) - fix(cmd): ensure proper working of mirror (https://github.com/zitadel/zitadel/pull/9509) - feat(Authz): system user support for permission check v2 (https://github.com/zitadel/zitadel/pull/9640) - chore(license): change from Apache to AGPL (https://github.com/zitadel/zitadel/pull/9597) - feat(console): list v2 sessions (https://github.com/zitadel/zitadel/pull/9539) - fix(console): add loginV2 feature flag (https://github.com/zitadel/zitadel/pull/9682) - fix(feature flags): allow reading "own" flags (https://github.com/zitadel/zitadel/pull/9649) - feat(console): add Actions V2 UI (https://github.com/zitadel/zitadel/pull/9591) BREAKING CHANGE - feat(webkey): migrate to v2beta API (https://github.com/zitadel/zitadel/pull/9445) - chore!: remove CockroachDB Support (https://github.com/zitadel/zitadel/pull/9444) - feat(actions): migrate to v2beta API (https://github.com/zitadel/zitadel/pull/9489) --------- Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Silvan <27845747+adlerhurst@users.noreply.github.com> Co-authored-by: Ramon <mail@conblem.me> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Kenta Yamaguchi <56732734+KEY60228@users.noreply.github.com> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Livio Spring <livio@zitadel.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Iraq <66622793+kkrime@users.noreply.github.com> Co-authored-by: Florian Forster <florian@zitadel.com> Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Max Peintner <peintnerm@gmail.com>
2025-04-02 16:53:06 +02:00
pushPlaceholderFmt = "($%d, $%d, $%d, $%d, $%d, $%d, $%d, $%d, $%d, statement_timestamp(), EXTRACT(EPOCH FROM clock_timestamp()), $%d)"
// uniqueConstraintPlaceholderFmt defines the format of the unique constraint error returned from the database
chore!: Introduce ZITADEL v3 (#9645) This PR summarizes multiple changes specifically only available with ZITADEL v3: - feat: Web Keys management (https://github.com/zitadel/zitadel/pull/9526) - fix(cmd): ensure proper working of mirror (https://github.com/zitadel/zitadel/pull/9509) - feat(Authz): system user support for permission check v2 (https://github.com/zitadel/zitadel/pull/9640) - chore(license): change from Apache to AGPL (https://github.com/zitadel/zitadel/pull/9597) - feat(console): list v2 sessions (https://github.com/zitadel/zitadel/pull/9539) - fix(console): add loginV2 feature flag (https://github.com/zitadel/zitadel/pull/9682) - fix(feature flags): allow reading "own" flags (https://github.com/zitadel/zitadel/pull/9649) - feat(console): add Actions V2 UI (https://github.com/zitadel/zitadel/pull/9591) BREAKING CHANGE - feat(webkey): migrate to v2beta API (https://github.com/zitadel/zitadel/pull/9445) - chore!: remove CockroachDB Support (https://github.com/zitadel/zitadel/pull/9444) - feat(actions): migrate to v2beta API (https://github.com/zitadel/zitadel/pull/9489) --------- Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Silvan <27845747+adlerhurst@users.noreply.github.com> Co-authored-by: Ramon <mail@conblem.me> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Kenta Yamaguchi <56732734+KEY60228@users.noreply.github.com> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Livio Spring <livio@zitadel.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Iraq <66622793+kkrime@users.noreply.github.com> Co-authored-by: Florian Forster <florian@zitadel.com> Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Max Peintner <peintnerm@gmail.com>
2025-04-02 16:53:06 +02:00
uniqueConstraintPlaceholderFmt = "(%s, %s, %s)"
_ eventstore.Pusher = (*Eventstore)(nil)
)
type Eventstore struct {
client *database.DB
perf(actionsv2): execution target router (#10564) # Which Problems Are Solved The event execution system currently uses a projection handler that subscribes to and processes all events for all instances. This creates a high static cost because the system over-fetches event data, handling many events that are not needed by most instances. This inefficiency is also reflected in high "rows returned" metrics in the database. # How the Problems Are Solved Eliminate the use of a project handler. Instead, events for which "execution targets" are defined, are directly pushed to the queue by the eventstore. A Router is populated in the Instance object in the authz middleware. - By joining the execution targets to the instance, no additional queries are needed anymore. - As part of the instance object, execution targets are now cached as well. - Events are queued within the same transaction, giving transactional guarantees on delivery. - Uses the "insert many fast` variant of River. Multiple jobs are queued in a single round-trip to the database. - Fix compatibility with PostgreSQL 15 # Additional Changes - The signing key was stored as plain-text in the river job payload in the DB. This violated our [Secrets Storage](https://zitadel.com/docs/concepts/architecture/secrets#secrets-storage) principle. This change removed the field and only uses the encrypted version of the signing key. - Fixed the target ordering from descending to ascending. - Some minor linter warnings on the use of `io.WriteString()`. # Additional Context - Introduced in https://github.com/zitadel/zitadel/pull/9249 - Closes https://github.com/zitadel/zitadel/issues/10553 - Closes https://github.com/zitadel/zitadel/issues/9832 - Closes https://github.com/zitadel/zitadel/issues/10372 - Closes https://github.com/zitadel/zitadel/issues/10492 --------- Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> (cherry picked from commit a9ebc06c778e1f46e04ff2b56f8ec4f337375aec)
2025-09-01 08:21:10 +03:00
queue eventstore.ExecutionQueue
}
refactor(eventstore): move push logic to sql (#8816) # Which Problems Are Solved If many events are written to the same aggregate id it can happen that zitadel [starts to retry the push transaction](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/eventstore.go#L101) because [the locking behaviour](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/v3/sequence.go#L25) during push does compute the wrong sequence because newly committed events are not visible to the transaction. These events impact the current sequence. In cases with high command traffic on a single aggregate id this can have severe impact on general performance of zitadel. Because many connections of the `eventstore pusher` database pool are blocked by each other. # How the Problems Are Solved To improve the performance this locking mechanism was removed and the business logic of push is moved to sql functions which reduce network traffic and can be analyzed by the database before the actual push. For clients of the eventstore framework nothing changed. # Additional Changes - after a connection is established prefetches the newly added database types - `eventstore.BaseEvent` now returns the correct revision of the event # Additional Context - part of https://github.com/zitadel/zitadel/issues/8931 --------- Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Miguel Cabrerizo <30386061+doncicuto@users.noreply.github.com> Co-authored-by: Joakim Lodén <Loddan@users.noreply.github.com> Co-authored-by: Yxnt <Yxnt@users.noreply.github.com> Co-authored-by: Stefan Benz <stefan@caos.ch> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Zach H <zhirschtritt@gmail.com>
2024-12-04 14:51:40 +01:00
var (
textType = &pgtype.Type{
Name: "text",
OID: pgtype.TextOID,
Codec: pgtype.TextCodec{},
}
commandType = &pgtype.Type{
Codec: &pgtype.CompositeCodec{
Fields: []pgtype.CompositeCodecField{
{
Name: "instance_id",
Type: textType,
},
{
Name: "aggregate_type",
Type: textType,
},
{
Name: "aggregate_id",
Type: textType,
},
{
Name: "command_type",
Type: textType,
},
{
Name: "revision",
Type: &pgtype.Type{
Name: "int2",
OID: pgtype.Int2OID,
Codec: pgtype.Int2Codec{},
},
},
{
Name: "payload",
Type: &pgtype.Type{
Name: "jsonb",
OID: pgtype.JSONBOID,
Codec: &pgtype.JSONBCodec{
Marshal: json.Marshal,
Unmarshal: json.Unmarshal,
},
},
},
{
Name: "creator",
Type: textType,
},
{
Name: "owner",
Type: textType,
},
},
},
}
commandArrayCodec = &pgtype.Type{
Codec: &pgtype.ArrayCodec{
ElementType: commandType,
},
}
)
var typeMu sync.Mutex
func RegisterEventstoreTypes(ctx context.Context, conn *pgx.Conn) error {
// conn.TypeMap is not thread safe
typeMu.Lock()
defer typeMu.Unlock()
m := conn.TypeMap()
var cmd *command
if _, ok := m.TypeForValue(cmd); ok {
return nil
}
if commandType.OID == 0 || commandArrayCodec.OID == 0 {
err := conn.QueryRow(ctx, "select oid, typarray from pg_type where typname = $1 and typnamespace = (select oid from pg_namespace where nspname = $2)", "command", "eventstore").
Scan(&commandType.OID, &commandArrayCodec.OID)
if err != nil {
logging.WithError(err).Debug("failed to get oid for command type")
return nil
}
if commandType.OID == 0 || commandArrayCodec.OID == 0 {
logging.Debug("oid for command type not found")
return nil
}
}
m.RegisterTypes([]*pgtype.Type{
{
Name: "eventstore.command",
Codec: commandType.Codec,
OID: commandType.OID,
},
{
Name: "command",
Codec: commandType.Codec,
OID: commandType.OID,
},
{
Name: "eventstore._command",
Codec: commandArrayCodec.Codec,
OID: commandArrayCodec.OID,
},
{
Name: "_command",
Codec: commandArrayCodec.Codec,
OID: commandArrayCodec.OID,
},
})
dialect.RegisterDefaultPgTypeVariants[command](m, "eventstore.command", "eventstore._command")
dialect.RegisterDefaultPgTypeVariants[command](m, "command", "_command")
return nil
}
// Client implements the [eventstore.Pusher]
func (es *Eventstore) Client() *database.DB {
return es.client
}
perf(actionsv2): execution target router (#10564) # Which Problems Are Solved The event execution system currently uses a projection handler that subscribes to and processes all events for all instances. This creates a high static cost because the system over-fetches event data, handling many events that are not needed by most instances. This inefficiency is also reflected in high "rows returned" metrics in the database. # How the Problems Are Solved Eliminate the use of a project handler. Instead, events for which "execution targets" are defined, are directly pushed to the queue by the eventstore. A Router is populated in the Instance object in the authz middleware. - By joining the execution targets to the instance, no additional queries are needed anymore. - As part of the instance object, execution targets are now cached as well. - Events are queued within the same transaction, giving transactional guarantees on delivery. - Uses the "insert many fast` variant of River. Multiple jobs are queued in a single round-trip to the database. - Fix compatibility with PostgreSQL 15 # Additional Changes - The signing key was stored as plain-text in the river job payload in the DB. This violated our [Secrets Storage](https://zitadel.com/docs/concepts/architecture/secrets#secrets-storage) principle. This change removed the field and only uses the encrypted version of the signing key. - Fixed the target ordering from descending to ascending. - Some minor linter warnings on the use of `io.WriteString()`. # Additional Context - Introduced in https://github.com/zitadel/zitadel/pull/9249 - Closes https://github.com/zitadel/zitadel/issues/10553 - Closes https://github.com/zitadel/zitadel/issues/9832 - Closes https://github.com/zitadel/zitadel/issues/10372 - Closes https://github.com/zitadel/zitadel/issues/10492 --------- Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> (cherry picked from commit a9ebc06c778e1f46e04ff2b56f8ec4f337375aec)
2025-09-01 08:21:10 +03:00
func NewEventstore(client *database.DB, opts ...EventstoreOption) *Eventstore {
es := &Eventstore{
client: client,
}
for _, opt := range opts {
opt(es)
}
return es
}
func (es *Eventstore) Health(ctx context.Context) error {
return es.client.PingContext(ctx)
}
refactor(eventstore): move push logic to sql (#8816) # Which Problems Are Solved If many events are written to the same aggregate id it can happen that zitadel [starts to retry the push transaction](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/eventstore.go#L101) because [the locking behaviour](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/v3/sequence.go#L25) during push does compute the wrong sequence because newly committed events are not visible to the transaction. These events impact the current sequence. In cases with high command traffic on a single aggregate id this can have severe impact on general performance of zitadel. Because many connections of the `eventstore pusher` database pool are blocked by each other. # How the Problems Are Solved To improve the performance this locking mechanism was removed and the business logic of push is moved to sql functions which reduce network traffic and can be analyzed by the database before the actual push. For clients of the eventstore framework nothing changed. # Additional Changes - after a connection is established prefetches the newly added database types - `eventstore.BaseEvent` now returns the correct revision of the event # Additional Context - part of https://github.com/zitadel/zitadel/issues/8931 --------- Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Miguel Cabrerizo <30386061+doncicuto@users.noreply.github.com> Co-authored-by: Joakim Lodén <Loddan@users.noreply.github.com> Co-authored-by: Yxnt <Yxnt@users.noreply.github.com> Co-authored-by: Stefan Benz <stefan@caos.ch> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Zach H <zhirschtritt@gmail.com>
2024-12-04 14:51:40 +01:00
var errTypesNotFound = errors.New("types not found")
func CheckExecutionPlan(ctx context.Context, conn *sql.Conn) error {
return conn.Raw(func(driverConn any) error {
if _, ok := driverConn.(sqlmock.SqlmockCommon); ok {
return nil
}
conn, ok := driverConn.(*stdlib.Conn)
if !ok {
return errTypesNotFound
}
return RegisterEventstoreTypes(ctx, conn.Conn())
})
}
func (es *Eventstore) pushTx(ctx context.Context, client database.ContextQueryExecuter) (tx database.Tx, deferrable func(err error) error, err error) {
tx, ok := client.(database.Tx)
if ok {
return tx, nil, nil
}
beginner, ok := client.(database.Beginner)
if !ok {
beginner = es.client
}
tx, err = beginner.BeginTx(ctx, &sql.TxOptions{
chore!: Introduce ZITADEL v3 (#9645) This PR summarizes multiple changes specifically only available with ZITADEL v3: - feat: Web Keys management (https://github.com/zitadel/zitadel/pull/9526) - fix(cmd): ensure proper working of mirror (https://github.com/zitadel/zitadel/pull/9509) - feat(Authz): system user support for permission check v2 (https://github.com/zitadel/zitadel/pull/9640) - chore(license): change from Apache to AGPL (https://github.com/zitadel/zitadel/pull/9597) - feat(console): list v2 sessions (https://github.com/zitadel/zitadel/pull/9539) - fix(console): add loginV2 feature flag (https://github.com/zitadel/zitadel/pull/9682) - fix(feature flags): allow reading "own" flags (https://github.com/zitadel/zitadel/pull/9649) - feat(console): add Actions V2 UI (https://github.com/zitadel/zitadel/pull/9591) BREAKING CHANGE - feat(webkey): migrate to v2beta API (https://github.com/zitadel/zitadel/pull/9445) - chore!: remove CockroachDB Support (https://github.com/zitadel/zitadel/pull/9444) - feat(actions): migrate to v2beta API (https://github.com/zitadel/zitadel/pull/9489) --------- Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Silvan <27845747+adlerhurst@users.noreply.github.com> Co-authored-by: Ramon <mail@conblem.me> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Kenta Yamaguchi <56732734+KEY60228@users.noreply.github.com> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Livio Spring <livio@zitadel.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Iraq <66622793+kkrime@users.noreply.github.com> Co-authored-by: Florian Forster <florian@zitadel.com> Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Max Peintner <peintnerm@gmail.com>
2025-04-02 16:53:06 +02:00
Isolation: sql.LevelReadCommitted,
refactor(eventstore): move push logic to sql (#8816) # Which Problems Are Solved If many events are written to the same aggregate id it can happen that zitadel [starts to retry the push transaction](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/eventstore.go#L101) because [the locking behaviour](https://github.com/zitadel/zitadel/blob/48ffc902cc90237d693e7104fc742ee927478da7/internal/eventstore/v3/sequence.go#L25) during push does compute the wrong sequence because newly committed events are not visible to the transaction. These events impact the current sequence. In cases with high command traffic on a single aggregate id this can have severe impact on general performance of zitadel. Because many connections of the `eventstore pusher` database pool are blocked by each other. # How the Problems Are Solved To improve the performance this locking mechanism was removed and the business logic of push is moved to sql functions which reduce network traffic and can be analyzed by the database before the actual push. For clients of the eventstore framework nothing changed. # Additional Changes - after a connection is established prefetches the newly added database types - `eventstore.BaseEvent` now returns the correct revision of the event # Additional Context - part of https://github.com/zitadel/zitadel/issues/8931 --------- Co-authored-by: Tim Möhlmann <tim+github@zitadel.com> Co-authored-by: Livio Spring <livio.a@gmail.com> Co-authored-by: Max Peintner <max@caos.ch> Co-authored-by: Elio Bischof <elio@zitadel.com> Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> Co-authored-by: Miguel Cabrerizo <30386061+doncicuto@users.noreply.github.com> Co-authored-by: Joakim Lodén <Loddan@users.noreply.github.com> Co-authored-by: Yxnt <Yxnt@users.noreply.github.com> Co-authored-by: Stefan Benz <stefan@caos.ch> Co-authored-by: Harsha Reddy <harsha.reddy@klaviyo.com> Co-authored-by: Zach H <zhirschtritt@gmail.com>
2024-12-04 14:51:40 +01:00
ReadOnly: false,
})
if err != nil {
return nil, nil, err
}
return tx, func(err error) error { return database.CloseTransaction(tx, err) }, nil
}
perf(actionsv2): execution target router (#10564) # Which Problems Are Solved The event execution system currently uses a projection handler that subscribes to and processes all events for all instances. This creates a high static cost because the system over-fetches event data, handling many events that are not needed by most instances. This inefficiency is also reflected in high "rows returned" metrics in the database. # How the Problems Are Solved Eliminate the use of a project handler. Instead, events for which "execution targets" are defined, are directly pushed to the queue by the eventstore. A Router is populated in the Instance object in the authz middleware. - By joining the execution targets to the instance, no additional queries are needed anymore. - As part of the instance object, execution targets are now cached as well. - Events are queued within the same transaction, giving transactional guarantees on delivery. - Uses the "insert many fast` variant of River. Multiple jobs are queued in a single round-trip to the database. - Fix compatibility with PostgreSQL 15 # Additional Changes - The signing key was stored as plain-text in the river job payload in the DB. This violated our [Secrets Storage](https://zitadel.com/docs/concepts/architecture/secrets#secrets-storage) principle. This change removed the field and only uses the encrypted version of the signing key. - Fixed the target ordering from descending to ascending. - Some minor linter warnings on the use of `io.WriteString()`. # Additional Context - Introduced in https://github.com/zitadel/zitadel/pull/9249 - Closes https://github.com/zitadel/zitadel/issues/10553 - Closes https://github.com/zitadel/zitadel/issues/9832 - Closes https://github.com/zitadel/zitadel/issues/10372 - Closes https://github.com/zitadel/zitadel/issues/10492 --------- Co-authored-by: Stefan Benz <46600784+stebenz@users.noreply.github.com> (cherry picked from commit a9ebc06c778e1f46e04ff2b56f8ec4f337375aec)
2025-09-01 08:21:10 +03:00
type EventstoreOption func(*Eventstore)
func WithExecutionQueueOption(queue eventstore.ExecutionQueue) EventstoreOption {
return func(es *Eventstore) {
es.queue = queue
}
}