rasuvaeff / yii3-outbox
Transactional outbox pattern for Yii3
Requires (Dev)
- ergebnis/composer-normalize: ^2.51
- friendsofphp/php-cs-fixer: ^3.95
- infection/infection: ^0.33
- maglnet/composer-require-checker: ^4.17
- rasuvaeff/property-testing-testo: ^0.6
- rasuvaeff/rector-named-literals: ^1.0
- rector/rector: ^2.4
- roave/backward-compatibility-check: ^8.0
- testo/bridge-infection: 0.1.6
- testo/testo: ^0.10.25
- vimeo/psalm: ^6.16
This package is auto-updated.
Last update: 2026-08-20 16:13:44 UTC
README
Transactional outbox pattern implementation for Yii3. Provides a stateless core for reliably publishing messages with configurable retry policies.
Using an AI coding assistant? llms.txt has a compact API reference you can use. Projects using the llm/skills Composer plugin also get this package's agent skill synced into
.agents/skills/automatically on install.
Requirements
- PHP 8.3+
psr/clock^1.0psr/log^3.0
Installation
composer require rasuvaeff/yii3-outbox
Usage
Recording a message
use DateTimeImmutable; use Psr\Clock\ClockInterface; use Rasuvaeff\Yii3Outbox\InMemoryStorage; use Rasuvaeff\Yii3Outbox\Outbox; $clock = new class implements ClockInterface { public function now(): DateTimeImmutable { return new DateTimeImmutable(); } }; $outbox = new Outbox(storage: $storage, clock: $clock); $message = $outbox->record( type: 'order.created', payload: json_encode(['orderId' => 42]), aggregateId: 'order-42', );
The transactional guarantee
The pattern is only worth its name when the outbox write commits atomically with the business write it describes. The core cannot enforce this: it opens no transaction and knows nothing about your connection. Two obligations are therefore yours:
- Call
record()inside the same database transaction as the business write. - Use a storage that writes through the same connection as your business
tables —
rasuvaeff/yii3-outbox-dbtakes aConnectionInterfacefor exactly this reason.
$db->transaction(static function () use ($orders, $outbox, $order, $json): void { $orders->insert($order); $outbox->record( type: 'order.created', payload: $json, aggregateId: $order->id, ); });
Break either obligation and the guarantee is void: commit the order without the
message and the event is lost forever; commit the message without the order and
consumers observe an event that never happened. A storage backed by a different
database — or by a message broker — cannot provide this guarantee at all, and
InMemoryStorage is a test double, not a durable one.
Message ids
The message id is the primary key of the outbox table and, when messages are
exported to ClickHouse, the deduplication key of the ReplacingMergeTree.
Two ways to control it:
Pass the domain event's id — the right choice whenever the message mirrors an event that already has an identifier. Republishing the same event then cannot mint a second id, so the consumer has something stable to deduplicate on:
$outbox->record( type: 'order.created', payload: $json, aggregateId: 'order-42', id: $domainEvent->getId(), );
Bind a generator for messages that have no domain id. The default
RandomHexIdGenerator keeps the historical format (32 random hex characters);
a time-ordered id makes inserts append instead of scattering across InnoDB
pages and gives batches a stable order:
use Rasuvaeff\Yii3Outbox\MessageIdGeneratorInterface; // symfony/uid final readonly class Uuid7IdGenerator implements MessageIdGeneratorInterface { public function generate(): string { return \Symfony\Component\Uid\Uuid::v7()->toRfc4122(); } } // ramsey/uuid — equally monotonic within the same millisecond final readonly class RamseyUuid7IdGenerator implements MessageIdGeneratorInterface { public function generate(): string { return \Ramsey\Uuid\Uuid::uuid7()->toString(); } } $outbox = new Outbox(storage: $storage, clock: $clock, idGenerator: new Uuid7IdGenerator());
The package ships no UUID implementation and depends on no UUID library —
id is VARCHAR(255) in rasuvaeff/yii3-outbox-db, so any format fits and
the choice stays yours.
Implementing storage
claim() is the primitive the whole polling loop rests on — Processor calls
it, never findPending(). It must atomically move messages to Processing and
return them, so that two workers polling the same table never receive the same
message. findPending() is the read-only counterpart: safe for dashboards and
diagnostics, unsafe as a worker's fetch.
use Rasuvaeff\Yii3Outbox\StorageInterface; use Rasuvaeff\Yii3Outbox\OutboxMessage; final class DbStorage implements StorageInterface { public function save(OutboxMessage $message): void { // INSERT INTO outbox ... ON CONFLICT(id) DO UPDATE ... // Must run on the caller's connection so it commits with the business write. } public function claim(array $types = [], int $limit = 1000): array { // Atomically: SELECT ids of status = 'pending' [AND type IN (:types)] // LIMIT :limit FOR UPDATE SKIP LOCKED // then UPDATE outbox SET status = 'processing', claimed_by = :worker // WHERE id IN (...) — and return the claimed rows. // Every claimed message must end up markPublished(), markFailed(), // or save($msg->withStatus(Pending)); none may stay Processing. } public function findPending(array $types = [], int $limit = 1000): array { // SELECT * FROM outbox WHERE status = 'pending' // [AND type IN (:types)] LIMIT :limit -- empty $types = all types // Read-only: no atomicity, so two workers would both get the same rows. // For retry support, also return status = 'pending' with attempts > 0 } public function markPublished(OutboxMessage $message): void { // UPDATE outbox SET status = 'published' WHERE id = ? } public function markFailed(OutboxMessage $message): void { // UPDATE outbox SET status = 'failed' WHERE id = ? } public function getById(string $id): ?OutboxMessage { // SELECT * FROM outbox WHERE id = ? } }
Letting the storage apply the retry policy
claim() returns every Pending message, so Processor receives the ones
still waiting out their backoff and writes each of them straight back as
Pending. That is two writes per backing-off message per iteration, and each
one occupies a slot in batchSize that a message ready to go could have used —
with a large retry queue, fresh messages wait behind it.
A storage that can express the predicate in its own query language implements
RetryAwareStorageInterface, and Processor claims through it automatically:
use Rasuvaeff\Yii3Outbox\RetryAwareStorageInterface; final class DbStorage implements RetryAwareStorageInterface { public function claimReady( DateTimeImmutable $readyThreshold, int $maxAttempts, array $types = [], int $limit = 1000, ): array { // Same atomic claim as claim(), with one more condition: // AND (attempts >= :maxAttempts // OR last_attempt_at IS NULL // OR last_attempt_at <= :readyThreshold) } // ... the rest of StorageInterface unchanged }
The attempts >= :maxAttempts clause is not an optimisation and must not be
dropped. A message out of attempts cannot be retried, and the only thing left to
do with it is mark it Failed — which Processor can only do to a message the
storage handed it. Filter it out and nothing ever terminates it: it stays
Pending, invisible to an alert watching Failed, forever.
$readyThreshold comes from RetryPolicy::readyThreshold($now) — the delay is
the core's business, and an implementation must not reconstruct it. The
interface exists separately from StorageInterface because adding the parameter
to claim() itself would break every third-party implementation.
rasuvaeff/yii3-outbox-db implements it. A storage that does not is still
correct: Processor falls back to claim() and filters in PHP, exactly as
before.
Two consequences worth knowing before you alert on them:
ProcessingResult::$skippedcounts messages the batch claimed and discarded. Against a retry-aware storage there are none, so it reads0— the work it used to count is what this interface removes.- A message whose attempts are spent is terminated up to
delaySecondslater than before, since it now waits for a batch that includes it.
Implementing a publisher
use Rasuvaeff\Yii3Outbox\PublisherInterface; use Rasuvaeff\Yii3Outbox\OutboxMessage; use Rasuvaeff\Yii3Outbox\PublishException; final class RabbitPublisher implements PublisherInterface { public function publish(OutboxMessage $message): void { try { // publish to RabbitMQ, Kafka, etc. } catch (\Throwable $e) { throw new PublishException( message: $e->getMessage(), outboxMessage: $message, previous: $e, ); } } }
Processing the outbox
use Rasuvaeff\Yii3Outbox\Processor; use Rasuvaeff\Yii3Outbox\RetryPolicy; $processor = new Processor( storage: $storage, publisher: $publisher, retryPolicy: new RetryPolicy(maxAttempts: 3, delaySeconds: 60), clock: $clock, batchSize: 100, ); $result = $processor->process(); // $result->published — successfully published // $result->failed — publish failures and messages that ran out of attempts // $result->skipped — claimed but not yet ready for retry; always 0 against // a RetryAwareStorageInterface, which never claims them
Retry behaviour
When a publish fails:
- If attempts <
maxAttempts→ message staysPending, will be retried afterdelaySeconds - If attempts >=
maxAttempts→ message is markedFailed(terminal)
Every message a batch claims leaves Processing. A message that is claimed
with its attempts already spent — restored from a backup, or left behind by a
markFailed() that never reached the database — is marked Failed on sight
rather than saved back as Pending, which would make it circle claim → skip →
save forever with no alert on Failed ever firing.
A publisher that throws something other than PublishException is a bug, and
process() rethrows it: the failure is not silently retried as if it were a
delivery problem. Before the exception propagates, the current message is
persisted per the retry policy and every message the batch had claimed but not
yet attempted is released back to Pending (or failed, if it too was out of
attempts), so nothing is left in Processing for a human to find with raw SQL.
The same applies when the storage — not the publisher — is what fails. If
markPublished() throws after publish() succeeded, the message reached its
consumer and only the record of that did not: it goes back to Pending and a
later run publishes it again. This package delivers at least once, and the
message id is what a consumer deduplicates on; leaving the row Processing
instead would be a row nothing in this API can move. That branch logs
Outbox message was published but could not be marked published, not the
publisher warning — during a storage incident an operator should not be sent to
debug the publisher.
The release is best-effort by construction: it reaches for the same storage that
may be the reason the batch aborted, and a storage that is down cannot be told
anything. What it does guarantee is that its own failures are logged
(Failed to release a claimed outbox message) rather than thrown — the
exception you catch is always the one that aborted the batch, never a symptom
raised while reacting to it.
$policy = new RetryPolicy(maxAttempts: 3, delaySeconds: 60); $policy->shouldRetry($message); // bool — attempts remaining? $policy->isReadyForRetry($message, $now); // bool — delay elapsed? $policy->readyThreshold($now); // DateTimeImmutable — the same question // as a boundary a storage can filter on
Using InMemoryStorage for tests
use Rasuvaeff\Yii3Outbox\InMemoryStorage; $storage = new InMemoryStorage(); $storage->save($message); $pending = $storage->findPending(); $storage->count(); $storage->clear();
API reference
Outbox
| Method | Description |
|---|---|
__construct(storage, clock, idGenerator?) |
Main entry point; default generator = RandomHexIdGenerator |
record(type, payload, aggregateId?, id?) |
Create and persist message, returns OutboxMessage. id = the domain event's id; omitted → generator. Call inside the business transaction |
StorageInterface
| Method | Description |
|---|---|
save(message) |
Persist. Must commit with the business write — see The transactional guarantee |
claim(types = [], limit = 1000) |
Atomically moves up to limit Pending messages to Processing and returns them. What Processor uses; safe for concurrent workers |
findPending(types = [], limit = 1000) |
Read-only listing of Pending messages. No atomicity — for dashboards, not for workers |
markPublished(message) |
Terminal success |
markFailed(message) |
Terminal failure |
getById(id) |
?OutboxMessage |
types filters by message type (empty = all), which is how several consumers
share one outbox. Since claim() hands a message to exactly one caller, the
type sets of independent consumers must not overlap — otherwise each message
reaches only whichever worker claimed it first.
RetryAwareStorageInterface
Extends StorageInterface. Optional: implement it when the backend can apply
the retry policy inside the claim itself — see
Letting the storage apply the retry policy.
| Method | Description |
|---|---|
claimReady(readyThreshold, maxAttempts, types = [], limit = 1000) |
Like claim(), but skips messages still waiting for their next attempt. Takes a message when it has never been attempted, was last attempted at or before readyThreshold, or has already spent maxAttempts attempts |
OutboxMessage
| Method | Description |
|---|---|
create(type, payload, aggregateId?, createdAt?, id?) |
Factory; id omitted → 32-char hex |
getId() |
Message ID (32-char hex) |
getType() |
Message type |
getPayload() |
Raw payload string |
getStatus() |
OutboxStatus enum |
getCreatedAt() |
DateTimeImmutable |
getAttempts() |
Number of publish attempts |
getLastAttemptAt() |
?DateTimeImmutable |
getAggregateId() |
?string |
withStatus(status) |
Returns new instance with status |
withAttempt(at) |
Returns new instance with incremented attempts and timestamp |
MessageIdGeneratorInterface
| Implementation | Produces |
|---|---|
RandomHexIdGenerator (default) |
32 hex characters, 128 random bits |
| your own | anything non-empty; id is VARCHAR(255) in the DB adapter |
OutboxStatus
| Case | Value | Meaning |
|---|---|---|
Pending |
'pending' |
Awaiting publication, including retries with attempts > 0 |
Processing |
'processing' |
Claimed by a worker; no other worker may take it |
Published |
'published' |
Terminal success |
Failed |
'failed' |
Terminal failure, retries exhausted |
RetryPolicy
| Method | Description |
|---|---|
__construct(maxAttempts, delaySeconds) |
Default: 3 attempts, 60s delay |
shouldRetry(message) |
Checks attempt count |
isReadyForRetry(message, now) |
Checks attempts + delay elapsed |
readyThreshold(now) |
now - delaySeconds: the same check as a boundary a storage can filter on. Feeds RetryAwareStorageInterface::claimReady() |
Processor
| Method | Description |
|---|---|
__construct(storage, publisher, retryPolicy, clock, batchSize, logger) |
Default batch: 100 |
process() |
Returns ProcessingResult |
ProcessingResult
| Property/Method | Description |
|---|---|
$published |
Count of successfully published messages |
$failed |
Count of publish failures this run, plus messages claimed with no attempts left |
$skipped |
Count of claimed messages not ready for retry. 0 against a RetryAwareStorageInterface, which never claims them |
total() |
Sum of all counters |
Serializer
Serializer implements SerializerInterface — the extension point a storage
backend or transport uses to move a message across a boundary as a string. Swap
in your own implementation for a different wire format; the interface is the
contract, Serializer is the JSON default.
| Method | Description |
|---|---|
serialize(message) |
Message to JSON |
deserialize(data) |
JSON to Message |
deserialize() rejects every malformed input with InvalidArgumentException —
missing fields, wrong types, an unknown status, an unparsable or empty
datetime. A caller catching "bad input" never has to also catch ValueError or
DateMalformedStringException from a field the parser forgot to guard.
Security
- Storage implementations must use parameterized queries for all user values.
- Message payload is stored as-is; validate before saving if needed.
Examples
See examples/ for complete usage examples.
Development
make install
make build
make cs-fix
make test
make test-coverage
make mutation
make release-check
make test-coverage and make mutation bootstrap pcov inside the
composer:2 container because the base image has no coverage driver.
License
BSD-3-Clause. See LICENSE.md.