zitadel/internal/database/cockroach/crdb.go

243 lines
6.1 KiB
Go
Raw Normal View History

package cockroach
import (
"context"
"database/sql"
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 13:51:40 +00:00
"fmt"
"strconv"
"strings"
"time"
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 13:51:40 +00:00
"github.com/jackc/pgx/v5"
"github.com/jackc/pgx/v5/pgxpool"
"github.com/jackc/pgx/v5/stdlib"
"github.com/mitchellh/mapstructure"
"github.com/zitadel/logging"
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 13:51:40 +00:00
"github.com/zitadel/zitadel/internal/api/authz"
"github.com/zitadel/zitadel/internal/database/dialect"
)
func init() {
feat(cmd): mirror (#7004) # Which Problems Are Solved Adds the possibility to mirror an existing database to a new one. For that a new command was added `zitadel mirror`. Including it's subcommands for a more fine grained mirror of the data. Sub commands: * `zitadel mirror eventstore`: copies only events and their unique constraints * `zitadel mirror system`: mirrors the data of the `system`-schema * `zitadel mirror projections`: runs all projections * `zitadel mirror auth`: copies auth requests * `zitadel mirror verify`: counts the amount of rows in the source and destination database and prints the diff. The command requires one of the following flags: * `--system`: copies all instances of the system * `--instance <instance-id>`, `--instance <comma separated list of instance ids>`: copies only the defined instances The command is save to execute multiple times by adding the `--replace`-flag. This replaces currently existing data except of the `events`-table # Additional Changes A `--for-mirror`-flag was added to `zitadel setup` to prepare the new database. The flag skips the creation of the first instances and initial run of projections. It is now possible to skip the creation of the first instance during setup by setting `FirstInstance.Skip` to true in the steps configuration. # Additional info It is currently not possible to merge multiple databases. See https://github.com/zitadel/zitadel/issues/7964 for more details. It is currently not possible to use files. See https://github.com/zitadel/zitadel/issues/7966 for more information. closes https://github.com/zitadel/zitadel/issues/7586 closes https://github.com/zitadel/zitadel/issues/7486 ### Definition of Ready - [x] I am happy with the code - [x] Short description of the feature/issue is added in the pr description - [x] PR is linked to the corresponding user story - [x] Acceptance criteria are met - [x] All open todos and follow ups are defined in a new ticket and justified - [x] Deviations from the acceptance criteria and design are agreed with the PO and documented. - [x] No debug or dead code - [x] My code has no repetitions - [x] Critical parts are tested automatically - [ ] Where possible E2E tests are implemented - [x] Documentation/examples are up-to-date - [x] All non-functional requirements are met - [x] Functionality of the acceptance criteria is checked manually on the dev system. --------- Co-authored-by: Livio Spring <livio.a@gmail.com>
2024-05-30 09:35:30 +00:00
config := new(Config)
dialect.Register(config, config, true)
}
const (
sslDisabledMode = "disable"
sslRequireMode = "require"
sslAllowMode = "allow"
sslPreferMode = "prefer"
)
type Config struct {
Host string
Port uint16
Database string
MaxOpenConns uint32
MaxIdleConns uint32
MaxConnLifetime time.Duration
MaxConnIdleTime time.Duration
User User
Admin AdminUser
// Additional options to be appended as options=<Options>
// The value will be taken as is. Multiple options are space separated.
Options string
}
func (c *Config) MatchName(name string) bool {
for _, key := range []string{"crdb", "cockroach"} {
if strings.TrimSpace(strings.ToLower(name)) == key {
return true
}
}
return false
}
feat(cmd): mirror (#7004) # Which Problems Are Solved Adds the possibility to mirror an existing database to a new one. For that a new command was added `zitadel mirror`. Including it's subcommands for a more fine grained mirror of the data. Sub commands: * `zitadel mirror eventstore`: copies only events and their unique constraints * `zitadel mirror system`: mirrors the data of the `system`-schema * `zitadel mirror projections`: runs all projections * `zitadel mirror auth`: copies auth requests * `zitadel mirror verify`: counts the amount of rows in the source and destination database and prints the diff. The command requires one of the following flags: * `--system`: copies all instances of the system * `--instance <instance-id>`, `--instance <comma separated list of instance ids>`: copies only the defined instances The command is save to execute multiple times by adding the `--replace`-flag. This replaces currently existing data except of the `events`-table # Additional Changes A `--for-mirror`-flag was added to `zitadel setup` to prepare the new database. The flag skips the creation of the first instances and initial run of projections. It is now possible to skip the creation of the first instance during setup by setting `FirstInstance.Skip` to true in the steps configuration. # Additional info It is currently not possible to merge multiple databases. See https://github.com/zitadel/zitadel/issues/7964 for more details. It is currently not possible to use files. See https://github.com/zitadel/zitadel/issues/7966 for more information. closes https://github.com/zitadel/zitadel/issues/7586 closes https://github.com/zitadel/zitadel/issues/7486 ### Definition of Ready - [x] I am happy with the code - [x] Short description of the feature/issue is added in the pr description - [x] PR is linked to the corresponding user story - [x] Acceptance criteria are met - [x] All open todos and follow ups are defined in a new ticket and justified - [x] Deviations from the acceptance criteria and design are agreed with the PO and documented. - [x] No debug or dead code - [x] My code has no repetitions - [x] Critical parts are tested automatically - [ ] Where possible E2E tests are implemented - [x] Documentation/examples are up-to-date - [x] All non-functional requirements are met - [x] Functionality of the acceptance criteria is checked manually on the dev system. --------- Co-authored-by: Livio Spring <livio.a@gmail.com>
2024-05-30 09:35:30 +00:00
func (_ *Config) Decode(configs []interface{}) (dialect.Connector, error) {
connector := new(Config)
decoder, err := mapstructure.NewDecoder(&mapstructure.DecoderConfig{
DecodeHook: mapstructure.StringToTimeDurationHookFunc(),
WeaklyTypedInput: true,
feat(cmd): mirror (#7004) # Which Problems Are Solved Adds the possibility to mirror an existing database to a new one. For that a new command was added `zitadel mirror`. Including it's subcommands for a more fine grained mirror of the data. Sub commands: * `zitadel mirror eventstore`: copies only events and their unique constraints * `zitadel mirror system`: mirrors the data of the `system`-schema * `zitadel mirror projections`: runs all projections * `zitadel mirror auth`: copies auth requests * `zitadel mirror verify`: counts the amount of rows in the source and destination database and prints the diff. The command requires one of the following flags: * `--system`: copies all instances of the system * `--instance <instance-id>`, `--instance <comma separated list of instance ids>`: copies only the defined instances The command is save to execute multiple times by adding the `--replace`-flag. This replaces currently existing data except of the `events`-table # Additional Changes A `--for-mirror`-flag was added to `zitadel setup` to prepare the new database. The flag skips the creation of the first instances and initial run of projections. It is now possible to skip the creation of the first instance during setup by setting `FirstInstance.Skip` to true in the steps configuration. # Additional info It is currently not possible to merge multiple databases. See https://github.com/zitadel/zitadel/issues/7964 for more details. It is currently not possible to use files. See https://github.com/zitadel/zitadel/issues/7966 for more information. closes https://github.com/zitadel/zitadel/issues/7586 closes https://github.com/zitadel/zitadel/issues/7486 ### Definition of Ready - [x] I am happy with the code - [x] Short description of the feature/issue is added in the pr description - [x] PR is linked to the corresponding user story - [x] Acceptance criteria are met - [x] All open todos and follow ups are defined in a new ticket and justified - [x] Deviations from the acceptance criteria and design are agreed with the PO and documented. - [x] No debug or dead code - [x] My code has no repetitions - [x] Critical parts are tested automatically - [ ] Where possible E2E tests are implemented - [x] Documentation/examples are up-to-date - [x] All non-functional requirements are met - [x] Functionality of the acceptance criteria is checked manually on the dev system. --------- Co-authored-by: Livio Spring <livio.a@gmail.com>
2024-05-30 09:35:30 +00:00
Result: connector,
})
if err != nil {
return nil, err
}
for _, config := range configs {
if err = decoder.Decode(config); err != nil {
return nil, err
}
}
feat(cmd): mirror (#7004) # Which Problems Are Solved Adds the possibility to mirror an existing database to a new one. For that a new command was added `zitadel mirror`. Including it's subcommands for a more fine grained mirror of the data. Sub commands: * `zitadel mirror eventstore`: copies only events and their unique constraints * `zitadel mirror system`: mirrors the data of the `system`-schema * `zitadel mirror projections`: runs all projections * `zitadel mirror auth`: copies auth requests * `zitadel mirror verify`: counts the amount of rows in the source and destination database and prints the diff. The command requires one of the following flags: * `--system`: copies all instances of the system * `--instance <instance-id>`, `--instance <comma separated list of instance ids>`: copies only the defined instances The command is save to execute multiple times by adding the `--replace`-flag. This replaces currently existing data except of the `events`-table # Additional Changes A `--for-mirror`-flag was added to `zitadel setup` to prepare the new database. The flag skips the creation of the first instances and initial run of projections. It is now possible to skip the creation of the first instance during setup by setting `FirstInstance.Skip` to true in the steps configuration. # Additional info It is currently not possible to merge multiple databases. See https://github.com/zitadel/zitadel/issues/7964 for more details. It is currently not possible to use files. See https://github.com/zitadel/zitadel/issues/7966 for more information. closes https://github.com/zitadel/zitadel/issues/7586 closes https://github.com/zitadel/zitadel/issues/7486 ### Definition of Ready - [x] I am happy with the code - [x] Short description of the feature/issue is added in the pr description - [x] PR is linked to the corresponding user story - [x] Acceptance criteria are met - [x] All open todos and follow ups are defined in a new ticket and justified - [x] Deviations from the acceptance criteria and design are agreed with the PO and documented. - [x] No debug or dead code - [x] My code has no repetitions - [x] Critical parts are tested automatically - [ ] Where possible E2E tests are implemented - [x] Documentation/examples are up-to-date - [x] All non-functional requirements are met - [x] Functionality of the acceptance criteria is checked manually on the dev system. --------- Co-authored-by: Livio Spring <livio.a@gmail.com>
2024-05-30 09:35:30 +00:00
return connector, nil
}
func (c *Config) Connect(useAdmin bool, pusherRatio, spoolerRatio float64, purpose dialect.DBPurpose) (*sql.DB, *pgxpool.Pool, error) {
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 13:51:40 +00:00
dialect.RegisterAfterConnect(func(ctx context.Context, c *pgx.Conn) error {
// CockroachDB by default does not allow multiple modifications of the same table using ON CONFLICT
// This is needed to fill the fields table of the eventstore during eventstore.Push.
_, err := c.Exec(ctx, "SET enable_multiple_modifications_of_table = on")
return err
})
connConfig, err := dialect.NewConnectionConfig(c.MaxOpenConns, c.MaxIdleConns, pusherRatio, spoolerRatio, purpose)
if err != nil {
return nil, nil, err
}
config, err := pgxpool.ParseConfig(c.String(useAdmin, purpose.AppName()))
if err != nil {
return nil, nil, err
}
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 13:51:40 +00:00
if len(connConfig.AfterConnect) > 0 {
config.AfterConnect = func(ctx context.Context, conn *pgx.Conn) error {
for _, f := range connConfig.AfterConnect {
if err := f(ctx, conn); err != nil {
return err
}
}
return nil
}
}
// For the pusher we set the app name with the instance ID
if purpose == dialect.DBPurposeEventPusher {
config.BeforeAcquire = func(ctx context.Context, conn *pgx.Conn) bool {
return setAppNameWithID(ctx, conn, purpose, authz.GetInstance(ctx).InstanceID())
}
config.AfterRelease = func(conn *pgx.Conn) bool {
ctx, cancel := context.WithTimeout(context.Background(), time.Second)
defer cancel()
return setAppNameWithID(ctx, conn, purpose, "IDLE")
}
}
if connConfig.MaxOpenConns != 0 {
config.MaxConns = int32(connConfig.MaxOpenConns)
}
config.MaxConnLifetime = c.MaxConnLifetime
config.MaxConnIdleTime = c.MaxConnIdleTime
pool, err := pgxpool.NewWithConfig(context.Background(), config)
if err != nil {
return nil, nil, err
}
if err := pool.Ping(context.Background()); err != nil {
return nil, nil, err
}
return stdlib.OpenDBFromPool(pool), pool, nil
}
func (c *Config) DatabaseName() string {
return c.Database
}
func (c *Config) Username() string {
return c.User.Username
}
func (c *Config) Password() string {
return c.User.Password
}
func (c *Config) Type() string {
return "cockroach"
}
func (c *Config) Timetravel(d time.Duration) string {
return ""
}
type User struct {
Username string
Password string
SSL SSL
}
type AdminUser struct {
// ExistingDatabase is the database to connect to before the ZITADEL database exists
ExistingDatabase string
User `mapstructure:",squash"`
}
type SSL struct {
// type of connection security
Mode string
// RootCert Path to the CA certificate
RootCert string
// Cert Path to the client certificate
Cert string
// Key Path to the client private key
Key string
}
func (c *Config) checkSSL(user User) {
if user.SSL.Mode == sslDisabledMode || user.SSL.Mode == "" {
user.SSL = SSL{Mode: sslDisabledMode}
return
}
if user.SSL.Mode == sslRequireMode || user.SSL.Mode == sslAllowMode || user.SSL.Mode == sslPreferMode {
return
}
if user.SSL.RootCert == "" {
logging.WithFields(
"cert set", user.SSL.Cert != "",
"key set", user.SSL.Key != "",
"rootCert set", user.SSL.RootCert != "",
).Fatal("at least ssl root cert has to be set")
}
}
func (c Config) String(useAdmin bool, appName string) string {
user := c.User
if useAdmin {
user = c.Admin.User
}
c.checkSSL(user)
fields := []string{
"host=" + c.Host,
"port=" + strconv.Itoa(int(c.Port)),
"user=" + user.Username,
"dbname=" + c.Database,
"application_name=" + appName,
"sslmode=" + user.SSL.Mode,
}
if c.Options != "" {
fields = append(fields, "options="+c.Options)
}
if !useAdmin {
fields = append(fields, "dbname="+c.Database)
} else if c.Admin.ExistingDatabase != "" {
fields = append(fields, "dbname="+c.Admin.ExistingDatabase)
}
if user.Password != "" {
fields = append(fields, "password="+user.Password)
}
if user.SSL.Mode != sslDisabledMode {
fields = append(fields, "sslrootcert="+user.SSL.RootCert)
if user.SSL.Cert != "" {
fields = append(fields, "sslcert="+user.SSL.Cert)
}
if user.SSL.Key != "" {
fields = append(fields, "sslkey="+user.SSL.Key)
}
}
return strings.Join(fields, " ")
}
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 13:51:40 +00:00
func setAppNameWithID(ctx context.Context, conn *pgx.Conn, purpose dialect.DBPurpose, id string) bool {
// needs to be set like this because psql complains about parameters in the SET statement
query := fmt.Sprintf("SET application_name = '%s_%s'", purpose.AppName(), id)
_, err := conn.Exec(ctx, query)
logging.OnError(err).Warn("failed to set application name")
return err == nil
}