Skip to content
bunqueue bunqueue
Get started
Get started

Client SDKs for Node.js, Deno, Python and Cloudflare Workers

guide · sdks

One protocol, every runtime.

Official client SDKs for Node.js, Deno, Bun, Python and Cloudflare Workers. Each one speaks the native TCP protocol, msgpack and pipelined, with full feature parity with the built-in Bun client.

100 e2e scenarios each on Node, Deno, Bun66 e2e scenarios in Python16 scenarios inside workerd1 versioned wire protocol

The design is simple: the server owns every queue semantic, retries with backoff, priorities, scheduling, stall detection, the dead letter queue. Your applications only add jobs and process them. A Next.js API written in TypeScript can enqueue work that a Python service consumes, both against the same queue, with no shared runtime and no translation layer.

Supported platforms

Section titled “Supported platforms”
PlatformPackageCurrentDistribution
Node.js ≥ 20, Bun, Deno ≥ 2, Cloudflare Workersbunqueue-client0.1.6npm · source
Python ≥ 3.9bunqueue-client0.1.1PyPI upcoming, installable from the repository today
Bun, embedded in process, no serverbunqueue—Quick Start

Both SDKs are versioned independently from the server and follow semantic versioning. The wire protocol is versioned too (protocolVersion: 1, negotiated via Hello), so a client keeps working across server upgrades.

Every claim on this page is verified against a real bunqueue server, on every supported platform: 100 end-to-end scenarios each on Node.js, Deno and Bun, 16 inside workerd (the actual Cloudflare Workers runtime), and 66 in Python, enforced in CI on every change (runtime matrix: Bun, Node and Deno for TypeScript; Python 3.10 and 3.12), with npm releases published with build provenance and gated on the full e2e suite.

Getting started

Section titled “Getting started”

1. Run the server

Section titled “1. Run the server”

The server is the only component that requires Bun.

Terminal window
bunx bunqueue start
# or, with Docker and persistent data:
# docker run -d -p 6789:6789 -p 6790:6790 -v bunqueue-data:/app/data ghcr.io/egeominotti/bunqueue:latest

Port 6789 serves the TCP protocol used by the SDKs, port 6790 serves the HTTP API with /health and /metrics. Additional deployment options are covered in Running the Server.

2. Install the client

Section titled “2. Install the client”
Terminal window
npm install bunqueue-client

3. Produce jobs

Section titled “3. Produce jobs”
import { Queue } from 'bunqueue-client';


const queue = new Queue('emails');


await queue.add('welcome', { to: 'user@example.com' });

4. Process jobs

Section titled “4. Process jobs”
import { Worker } from 'bunqueue-client';


const worker = new Worker('emails', async (job) => {
  await sendEmail(job.data.to);
  return { sent: true };
});


worker.on('completed', (job) => console.log('done:', job.id));
worker.on('error', (err) => console.error(err)); // always attach (see Worker semantics)

Run either file with the runtime you already use:

Terminal window
node --experimental-strip-types app.ts    # Node 22 or later
bun app.ts                                # Bun
deno run -A app.ts                        # Deno 2 or later
python app.py                             # Python

Producer and worker are usually separate services, often in different languages, a Next.js API adds jobs, a Python service processes them, against the same queue and the same protocol. Constructors default to host: 'localhost' and port: 6789, so no options are needed for a local setup.

Protocol and architecture

Section titled “Protocol and architecture”

Understanding four facts about the transport explains most SDK behavior:

  1. Framing, every message is a 4-byte big-endian length prefix followed by a standard msgpack map. Maximum frame size is 64 MB; maximum job payload is 10 MB.
  2. Pipelining, every request carries a reqId the server echoes back, so many commands are in flight on one socket concurrently. A single connection is usually all a service needs.
  3. Authentication-first, when a token is configured, Auth is guaranteed to be the first frame on every (re)connection. The TypeScript client enforces this through synchronous write ordering; the Python client through a connection lock (safe under free-threaded concurrency).
  4. The job name travels inside data, add('welcome', {to}) is encoded as data: { name: 'welcome', to }. Non-object payloads are wrapped as { name, payload }.

Connection options

Section titled “Connection options”
const queue = new Queue('emails', {
  host: 'queue.example.com',
  port: 6789,
  token: process.env.BUNQUEUE_TOKEN,
  tls: true,               // or { caFile } or { rejectUnauthorized: false }
  commandTimeoutMs: 10000, // default
  maxInFlight: 10_000,     // backpressure bound; 0 = unbounded (default)
  poolSize: 4,             // producer-side connection pool; default 1
});

The TCP connect timeout is internal in both SDKs (5 seconds) and is not configurable through the constructor; the command timeout above governs how long each in-flight command waits for a response.

Connection resilience

Section titled “Connection resilience”

Both SDKs implement the same production-grade recovery model, verified by dedicated failure-injection tests:

MechanismBehavior
Lazy reconnectA lost connection reconnects transparently on the next call; workers re-register automatically after every reconnect (generation counter)
Reconnect backoffExponential 500 ms → 5 s with a fast-fail window, so a producer calling add() against a downed server does not pay the full connect timeout on every call
TCP keepaliveEnabled at ~15 s idle: a half-open link (cloud NAT / load-balancer idle drop with no FIN/RST) is detected in seconds instead of the OS-level ~15 minutes
Timeout-driven teardownThree consecutive command timeouts tear the socket down so the next call reconnects, generation-guarded, so a stale timeout can never abort a freshly reconnected socket
Backpressure (TypeScript)With maxInFlight set, callers park until an in-flight slot frees instead of growing memory without bound under a fast producer / slow server
Auth orderingNo command can race ahead of the Auth frame on a reconnect, a dropped ACK caused by that race would otherwise resurface as a silent duplicate execution

Producing jobs

Section titled “Producing jobs”

Job options

Section titled “Job options”

Every option is validated server-side; identical semantics in both SDKs (Python uses snake_case: attempts, job_id, remove_on_complete, …).

OptionTypeDefaultNotes
prioritynumber0Higher runs sooner; −1 000 000 … 1 000 000
delayms0Up to 1 year
attemptsnumber3Max attempts including the first; up to 1000
backoffnumber | { type, delay }1000type: 'fixed' or 'exponential'; delay up to 1 day
ttlms—Expires the job if not processed in time
timeoutms—Per-job processing timeout, up to 1 day
jobIdstring—Custom id; idempotent, re-adding an unfinished id is a no-op
deduplication{ id, ttl?, extend?, replace? }—Dedup window keyed on id
debounce{ id, ttl? }—Collapses bursts to the last job
dependsOnstring[]—Job ids that must complete first
parentId / childrenIdsstring / string[]—Flow relationships (usually set via FlowProducer)
tags / groupIdstring[] / string—Metadata; groupId also scopes group rate limits
lifobooleanfalseOrders only among LIFO jobs; mixed queues stay FIFO
removeOnComplete / removeOnFailbooleanfalseDrop the job record at the terminal state
durablebooleanfalseBypass the server write buffer, fsync before the ACK returns
repeat{ every } or { pattern, tz? } + limit?—Repeatable jobs (see Cron)
stallTimeoutms—Per-job stall detection override
stackTraceLimit / keepLogs / sizeLimitnumber—Failure stack cap, retained log lines, payload cap
await queue.add('report', data, { priority: 10, delay: 5000, attempts: 5 });
await queue.add('charge', payment, { jobId: `order-${orderId}`, durable: true });

Idempotency and bulk

Section titled “Idempotency and bulk”

jobId makes add idempotent: re-adding an id whose job is still unfinished (waiting, active, waiting-children) returns the existing job instead of creating a duplicate. This holds for addBulk too, each bulk entry’s jobId is preserved on the wire, so an idempotent batch ingest can be re-run safely after a crash:

await queue.addBulk(
  orders.map((o) => ({ name: 'ingest', data: o, opts: { jobId: `order-${o.id}` } }))
);

Producer throughput

Section titled “Producer throughput”

For high-volume producers the TypeScript SDK can fan commands across a round-robin connection pool, one line, no API change:

const queue = new Queue('ingest', { poolSize: 4 });

The pool is producer-side only: workers intentionally keep a single connection, because job leases (lock tokens) and worker registration are per-connection server state.

Processing jobs

Section titled “Processing jobs”

Worker options

Section titled “Worker options”
OptionDefaultNotes
concurrency4Jobs processed in parallel
batchSize10Jobs fetched per PULLB, capped by free slots
pollTimeoutMs5000Server-side long-poll; max 30 000
lockTtlMs30 000Job lease TTL
heartbeatIntervalS10Worker + per-job lock heartbeats
ackBatchoffOpt-in ACK batching (below)
autoruntrueStart the pull loop at construction

Lease model

Section titled “Lease model”

A pulled job carries a lock token. The worker heartbeats every active job’s lock on the heartbeat interval, so a job that legitimately runs longer than the lock TTL survives. If the worker dies, the lease expires and the server requeues the job (or moves it to the DLQ once maxStalls is exceeded), at-least-once delivery, so make handlers idempotent.

Failures

Section titled “Failures”
import { UnrecoverableError } from 'bunqueue-client';


const worker = new Worker('emails', async (job) => {
  if (!isValid(job.data)) {
    throw new UnrecoverableError('malformed payload'); // skip retries → DLQ
  }
  return await send(job.data);
});

A thrown error fails the job with its message and the leading stack lines (persisted server-side, capped by stackTraceLimit); the server applies the retry/backoff policy and eventually the dead letter queue. UnrecoverableError bypasses retries entirely.

Worker events

Section titled “Worker events”

ready, active, completed, failed, progress, drained, cancelled, error, closed.

Always attach an error listener. Per Node EventEmitter semantics an unhandled error event throws. Both workers free each job’s concurrency slot before emitting, and the Python worker additionally swallows listener exceptions, so a throwing listener cannot leak a slot or degrade throughput in either SDK. The error itself is yours to observe.

ACK batching (high volume)

Section titled “ACK batching (high volume)”

Opt-in: coalesce completed-job acknowledgements into ACKB round-trips.

const worker = new Worker('ingest', process, {
  concurrency: 32,
  ackBatch: { enabled: true, maxSize: 50, maxDelayMs: 5 },
});

Semantics are strict: a job stays active, its lock still heartbeated, until the server confirms the batch; the batch is flushed on close(); every job settles exactly once even if an event listener throws. Defaults are off, so nothing changes unless you enable it.

Graceful shutdown

Section titled “Graceful shutdown”
await worker.close();        // stop pulling, flush batched ACKs, drain in-flight jobs
await worker.close(true);    // force: skip the in-flight drain

Query, control and operations

Section titled “Query, control and operations”

The full management surface is available in every SDK with identical semantics (Python: same methods in snake_case; the rare name that differs is noted inline).

AreaMethods
QuerygetJob, getJobByCustomId, getJobs + per-state helpers (getWaiting, getActive, getCompleted, getFailed, getDelayed, getPrioritized, getWaitingChildren), getJobState (Python name: get_state), getResult, getProgress, counts + counts-per-priority, getChildrenValues, job logs
WaitingwaitForJob(id, ttlMs), resolves with the result; throws on timeout (CommandTimeoutError) and on a failed job (CommandError), never silently returns undefined
Controlpause / resume / isPaused, drain, obliterate, clean, remove, discard, promoteJob(s), retryJob (failed → waiting), updateJobData, updateJobProgress, changeJobPriority, changeJobDelay, moveJobToWait/Delayed/Completed/Failed, extendJobLock
DLQgetDlq, retryDlq(id?), purgeDlq, retryJobs (BullMQ semantics: retries the whole DLQ by default, or pass count to retry only the first N entries, honored by server 2.8.29 and later; Python retry_dlq accepts count= too), DLQ configuration
SchedulersupsertJobScheduler (cron pattern or fixed interval, timezone, skipIfNoWorker, preventOverlap, per-scheduler job options), removeJobScheduler, getJobScheduler(s), addCron, every
Adminrate limiting, global concurrency, stall configuration, webhooks, getStats, getMetrics, listQueues, getWorkers

Two behaviors worth knowing:

  • Not-found is null/None, never an exception, getJob, getJobByCustomId and getJobScheduler map the server’s not-found response for you.
  • updateJobProgress requires an active job; the server rejects progress updates on waiting jobs by design.

Flows

Section titled “Flows”

Pipelines and parent/child trees with automatic ordering, children complete before their parent, results are readable from the parent:

import { FlowProducer } from 'bunqueue-client';


const flow = new FlowProducer();


// Sequential pipeline
await flow.addChain([
  { name: 'extract', queueName: 'etl' },
  { name: 'transform', queueName: 'etl' },
  { name: 'load', queueName: 'etl' },
]);


// Tree: parent waits for its children
const node = await flow.add({
  name: 'assemble', queueName: 'orders',
  children: [
    { name: 'reserve-stock', queueName: 'orders' },
    { name: 'charge-card', queueName: 'orders' },
  ],
});


// Fan-in: N parallel jobs converge into one
await flow.addBulkThen(parts, { name: 'merge', queueName: 'orders' });

Creation is transactional per call: if any push fails, every job already created by that call is cancelled (rollback). getFlow(id) reconstructs a tree from a root id, it returns null for a missing root, skips children that were since removed (removeOnComplete), and is cycle-safe, so it always returns the surviving partial tree instead of throwing.

Observability

Section titled “Observability”

The TypeScript SDK exposes a zero-dependency observability layer: you bring OpenTelemetry, Prometheus or your own logger, the SDK stays silent by default and consumer errors are isolated so a throwing sink can never break the transport.

import { Queue, consoleLogger, type TelemetryEvent } from 'bunqueue-client';


const queue = new Queue('emails', {
  logger: consoleLogger('info'),           // or any { debug, info, warn, error }
  onTelemetry: (e: TelemetryEvent) => {
    if (e.type === 'command') histogram.observe({ cmd: e.cmd }, e.durationMs);
    if (e.type === 'reconnect_scheduled') reconnects.inc();
  },
});


// Connection is an EventEmitter for imperative lifecycle hooks:
queue.connection.on('connect', (i) => log.info('link up', i));
queue.connection.on('disconnect', (i) => log.warn('link down', i));
queue.connection.on('reconnect_scheduled', (i) => log.warn('retrying', i));

TelemetryEvent is a typed union: command (per-command latency and outcome), command_timeout, connect, disconnect, reconnect_scheduled, auth, backpressure. For server-side metrics, scrape the HTTP /metrics endpoint (Prometheus format).

Simple Mode

Section titled “Simple Mode”

Bunqueue bundles a Queue and a Worker into a single object, a 1:1 port of the official client’s Simple Mode. It brings routes, onion middleware, in-process retry strategies, a circuit breaker, batch accumulation, event triggers, job TTL, priority aging, cooperative cancellation, and deduplication or debounce defaults.

import { Bunqueue } from 'bunqueue-client';


const app = new Bunqueue('notifications', {
  routes: {
    'send-email': async (job) => ({ sent: true }),
    'send-sms': async (job) => ({ sent: true }),
  },
  concurrency: 10,
  retry: { maxAttempts: 5, strategy: 'jitter' },
  circuitBreaker: { threshold: 5, resetTimeout: 30_000 },
});


app.use(async (job, next) => {
  console.time(job.name);
  const result = await next();
  console.timeEnd(job.name);
  return result;
});


await app.add('send-email', { to: 'alice@example.com' });
await app.cron('daily-digest', '0 9 * * *', { to: 'all' });

Scheduler job options are honored end to end: the template’s attempts, backoff, timeout, delay, remove_on_complete / remove_on_fail, priority and deduplication are applied to every cron-spawned job, in both SDKs.

One difference from the Bun original: embedded: true is not available, it requires the in-process Bun runtime. Connect to a server instead; the option raises a clear error if you try.

Cloudflare Workers

Section titled “Cloudflare Workers”

The same client runs inside Workers. Enable Node.js compatibility and add jobs directly from your fetch handlers:

wrangler.toml
compatibility_flags = ["nodejs_compat"]
compatibility_date = "2025-01-01"
import { Queue } from 'bunqueue-client';


export default {
  async fetch(req: Request, env: Env): Promise<Response> {
    const queue = new Queue('signups', { host: env.BQ_HOST, port: 6789, token: env.BQ_TOKEN, tls: true });
    const job = await queue.add('welcome', await req.json());
    queue.close();
    return Response.json({ queued: job.id });
  },
};

Workers are request-scoped, so there is no long-lived worker loop. Instead: produce from fetch handlers, and consume in batches from a Cron Trigger, pull, process, acknowledge, return. Both patterns (Simple Mode included) are covered by the SDK’s 16-scenario suite that runs inside workerd. Two requirements apply: the server must be reachable from the internet, and TLS needs a publicly trusted certificate.

Security

Section titled “Security”

Authentication uses server-side tokens; transport security uses native TLS.

const queue = new Queue('emails', {
  host: 'queue.example.com',
  port: 6789,
  token: process.env.BUNQUEUE_TOKEN,   // server started with AUTH_TOKENS=...
  tls: true,                            // or { caFile: './ca.pem' } for a custom CA
});

Certificate verification is on by default for every TLS connection, a wrong or missing CA rejects the connection instead of silently connecting; only an explicit rejectUnauthorized: false (TypeScript) or {"verify": False} (Python) opts into encryption-only mode for development. Start the server with AUTH_TOKENS to require authentication, and with TLS_CERT_FILE plus TLS_KEY_FILE for encrypted transport. Full hardening guidance lives in the deployment guide.

Errors and guarantees

Section titled “Errors and guarantees”

Both SDKs expose the same typed error hierarchy, so retry logic can branch precisely:

ErrorMeaning
ConnectionClosedErrorLink lost or server unreachable (in-flight commands reject; the next call reconnects)
CommandTimeoutErrorNo response within the command timeout, also thrown by waitForJob on timeout
CommandErrorThe server answered ok: false (validation error, not-found on write ops, failed job in waitForJob)
AuthErrorToken rejected during the connection handshake
UnrecoverableErrorThrown by your processor to skip retries and fail terminally

Delivery is at-least-once: a worker crash after processing but before the ACK means the job runs again, design handlers to be idempotent (the jobId option helps on the producing side). Two numeric-precision rules:

  • JavaScript numbers are IEEE 754 doubles, exact up to 2⁵³. Pass larger 64-bit identifiers (snowflake ids) as strings; never put BigInt in job data.
  • Python integers outside the int32 range are automatically encoded as float64 on the wire (exact up to 2⁵³, safe for millisecond timestamps); the same string rule applies to larger identifiers.

Quality assurance

Section titled “Quality assurance”

Every SDK release is validated against a real bunqueue server, spawned fresh for each run, on every supported runtime, locally and in CI on every change:

RuntimeVerification
Node.js100 e2e scenarios and 8 integration tests
Bun100 e2e scenarios and 8 integration tests
Deno100 e2e scenarios and 8 integration tests, identical test files
Cloudflare Workers16 scenarios executed inside workerd, full API surface and Simple Mode
Python66 e2e scenarios and 8 integration tests

Coverage includes payload limits, unicode integrity, pipelining under concurrency, server crash and restart with automatic reconnection, half-open-link recovery, auth ordering under concurrent commands, ACK-batch error paths (a failing batch settles every job exactly once and leaks no concurrency slot), lock survival for jobs that outlast the lock TTL, and a realistic multi-queue production scenario with zero-loss accounting. npm releases are published with build provenance and gated on the e2e suite.

Browsers are not supported, raw TCP sockets are unavailable there, so route through your own API instead.

Resources

Section titled “Resources”