Skip to content

Latest commit

 

History

History
1997 lines (1448 loc) · 73.3 KB

CHANGELOG.md

File metadata and controls

1997 lines (1448 loc) · 73.3 KB
Raw

CHANGELOG

🚀 v2023.1.0-rc.2 🚀

👀 New:

  • ✒️ Velox: Support for v2023 and v4 plugins. v2andv3` plugins are still supported.

🩹 Fixes:

  • 🐛 Service plugin: Fix deadlock on channel send operation when using remain_after_exit.
  • 🐛 Service plugin: Send SIGINT instead of SIGKILL to underlying processes with 5s timeout to let the process exit gracefully.
  • 🐛 Config plugin: Fix missing default env variable syntax parser, BUG, (thanks @benalf)

🚀 v2023.1.0-rc.1 🚀

👀 New:

  • ✒️ Server plugin: pass RR_VERSION env variable to the worker to check the current RR version. Version passed without the v prefix (e.g. 2023.1.0).
  • ✒️ Lock plugin: faster first call to acquire the lock.

🚀 v2023.1.0-beta.1 🚀

👀 New:

  • ✒️ Kafka plugin: Totally reworked Kafka plugin. Now it supports regexps for the topics, marked commits for the group consumers, and SASL authentication. Configuration reference: link.
  • ✒️ RPC plugin: RPC plugin would be available immediately before the worker initialization. That means, that PHP worker may use all RPC methods immediately.
  • ✒️ Endure v2 support (internal change).
  • ✒️ Bash script to download the latest RR archive. Later we're going to release a non archived binaries in addition to the regular archived releases. Sample of usage:
curl --proto '=https' --tlsv1.2 -sSf  https://raw.githubusercontent.com/roadrunner-server/roadrunner/master/download-latest.sh | sh
  • ✒️ RoadRunner composer metapackage: Removed the require section: PR, (thanks @roxblnfk)
  • ✒️ Lock plugin: New plugin to handle mutual access to the resource. PHP client is WIP with documentation.
  • ✒️ AMQP plugin: RR passes the Queue, Pipeline and Driver names to the PHP client in all modes including the consuming payloads from the other senders.
  • ✒️ AMQP plugin: consumer_id can now be set from the configuration, FR, (thanks @codercms)
  • ✒️ AMQP plugin: Starting from the v2023.1.0 RR would not accept the empty queue name, CH
  • ✒️ OTEL plugin: ️Support OpenTelemetry for the temporal, http, gRPC and Jobs plugins including all Jobs drivers.
  • ✒️ Config plugin: Configuration version updated to the version: '3'. ️
  • ✒️ Logger plugin: Now uses UTC timestamps CH, (thanks @cv65kr)
  • ✒️ gRPC plugin: Support user defined interceptors.
  • ✒️ Temporal plugin: Support user defined interceptors.

🩹 Fixes:

  • 🐛 HTTP plugin: Edge case where empty form value overwrites existing value, PR, (thanks @tungfinblox).
  • 🐛 AMQP plugin: Redial failed if user uses only consumer, PR, (thanks @iborysenko).
  • 🐛 RR CLI: ./rr jobs command panic if used without arguments, BUG, (thanks @embargo2710)
tags: roadrunner v2.12.3

v2.12.3 (16.02.2023)

🚀 v2.12.3 🚀

👀 New:

  • ✒️ Composer.json: add contributors, funds, project description: PR, (thanks @roxblnfk)

🧹 Chore:

  • 🧑‍🏭 Dependencies: update project dependencies.
  • 🧑‍🏭 Go: update Go to 1.20.
tags: roadrunner v2.12.2

v2.12.2 (12.01.2023)

🚀 v2.12.2 🚀

👀 New:

  • ✒️ AMQP plugin: Custom headers in AMQP driver, FR, (thanks @ykweb)

🩹 Fixes:

  • 🐛 Velox: Unable to build RoadRunner with custom velox configuration, BUG, (thanks @mprokocki)
  • 🐛 RR: JSON Schema - wrong type for service exec_timeout option, BUG, (thanks @Chi-teck)
  • 🐛 RR: Fix the description of the --silent flag, PR, (thanks @maximal)
tags: roadrunner v2.12.2

🚀 v2.12.2-alpha.1 🚀

👀 New:

  • ✏ AMQP plugin: pass headers to the Queue declaration command to create quorum, lazy, etc. queues and provide additional options supported by RabbitMQ, FR, (thanks @ykweb)

  • ✏ AMQP plugin: do not create a queue if the user does not consume it, FR, (thanks @Colomix)

🩹 Fixes:

  • 🐛 X-Sendfile middleware: make it work as expected (as the response header), BUG, (thanks @tux-rampage)

v2.12.1 (01.12.2022)

tags: roadrunner v2.12.1

🚀 v2.12.1 🚀

👀 New:

  • ✏ RR: Automatically set the GOMAXPROCS to match the container CPU quota.
  • ✏ AMQP plugin: implement Status to check the AMQP connection PR.
  • ✏ SQS plugin: prefetch option now works as expected. RR will not consume new JOBS when reaches prefetch limit, until already accepted messages are not ACK/NACK-ed.
  • ✏ JOBS(memory) plugin: prefetch option now works as expected (see SQS). You can now emulate FIFO in memory by setting the prefetch option to 1.

🩹 Fixes:

  • gRPC plugin: server options are applied only when TLS is set. Discussion.
  • AMQP plugin: fix a few typos in the configuration.

v2.12.0 (24.11.2022)

⚠️ websocket and broadcast plugins were replaced by the new centrifuge plugin.

⚠️ All plugins, sdk and api updated to v3. There are no breaking changes, we moved all Go code from the api to sdk.

👀 New:

  • ✏ All plugins: update to v3. This is done not because of some breaking change but because of the internal update.
  • ✏ RPC plugin: add new API to provide a running RR version and RR configuration in JSON format.
  • ✏ Metrics plugin: add new API to unregister previously registered collector. (thanks @butschster)
  • ✏ Server plugin: add new API to get statuses about the service and its child processes. (thanks @butschster)
  • ✏ App logger plugin: Application logger plugin, FR (thanks @wolfy-j) Docs: PHP-lib
  • ✏ AMQP plugin: new configuration options. FR, (thanks @andrey-tech)
jobs:
  pipelines:
    example:
      driver: amqp
      config:
        # Durable exchange
        #
        # Default: true
        exchange_durable: true

        # Auto-deleted exchange
        #
        # Default: false
        exchange_auto_deleted: false

        # Auto-deleted queue
        #
        # Default: false
        queue_auto_deleted: false
  • ✏ Workers pool (SDK): New option to control the reset_timeout. Note that the pool.Reset is protected by mutexes, meaning that if you have some requests already in the pool, you'll have to wait for these requests to be processed. The reset_timeout does not count this time.
pool:
  allocate_timeout: 10s
  reset_timeout: 10s
  destroy_timeout: 10s
  • ✏ Centrifugo plugin: New centrifugo plugin. Which is going to replace existing broadcast + websockets plugins. FR. Docs: PHP-lib

RoadRunner config:

version: "2.7"

centrifuge:
  # Centrifugo server proxy address (docs: https://centrifugal.dev/docs/server/proxy#grpc-proxy)
  #
  # Optional, default: tcp://127.0.0.1:30000
  proxy_address: "tcp://127.0.0.1:30000"

  # gRPC server API address (docs: https://centrifugal.dev/docs/server/server_api#grpc-api)
  #
  # Optional, default: tcp://127.0.0.1:30000. Centrifugo: `grpc_api` should be set to true and `grpc_port` should be the same as in the RR's config.
  grpc_api_address: tcp://127.0.0.1:30000

  # Use gRPC gzip compressor
  #
  # Optional, default: false
  use_compressor: true

  # Your application version
  #
  # Optional, default: v1.0.0
  version: "v1.0.0"

  # Your application name
  #
  # Optional, default: roadrunner
  name: "roadrunner"

  # TLS configuration
  #
  # Optional, default: null
  tls:
    # TLS key
    #
    # Required
    key: /path/to/key.pem

    # TLS certificate
    #
    # Required
    cert: /path/to/cert.pem


  # Workers pool settings. link: https://github.com/roadrunner-server/roadrunner/blob/master/.rr.yaml#L812
  #
  # Optional, default: null (see default values)
  pool: {}

🩹 Fixes:

  • 🐛 Headers middleware: Header size is too small, BUG (thanks @masterjus)
  • 🐛 gRPC plugin: Protobuf compiler plugin segfaults on import statements, BUG (thanks @phroggyy)
  • 🐛 Service plugin: Get services list via RPC, BUG (thanks @butschster)
  • 🐛 gRPC plugin: Remote protoc-gen-php-grpc plugin error, BUG (thanks @rapita)
  • 🐛 HTTP plugin: Fail to upload files when RR's permissions are different from worker's, BUG (thanks @egonbraun)

v2.11.4 (06.10.2022)

👀 New:

  • ✏ Temporal plugin: Support for the SearchAttributes. FR, (thanks @cv65kr).

Docs: link Samples: link

🔧 Maintenance:

  • roadrunner-temporal plugin updated to: 1.7.0
  • http plugin updated to: 2.23.5
  • sqs plugin updated to: 2.20.4
  • config plugin updated to: 2.16.5
  • grpc plugin updated to: 2.23.3
  • nats plugin updated to: 2.17.3
  • jobs plugin updated to: 2.18.4
  • server plugin updated to: 2.16.4
  • tcp plugin updated to: 2.15.4
  • websockets plugin updated to: 2.16.5
  • otel plugin updated to: 2.5.6
  • kafka plugin updated to: 2.2.3

v2.11.3 (29.09.2022)

👀 New:

  • ✏ [ALPHA] gRPC plugin: buf remote plugins support for the protoc-gen-php-grpc plugin. FR, (thanks @rauanmayemir)
  • ✏ Temporal plugin: mTLS support. FR, (thanks @seregazhuk)

Configuration sample:

temporal:
  address: 127.0.0.1:7233
  cache_size: 100000
  activities:
    num_workers: 4

  tls:
    key: client.key
    cert: client.pem
    root_ca: ca.cert
    client_auth_type: require_and_verify_client_cert
    server_name: "tls-sample"

🩹 Fixes:

  • 🐛 Config plugin: properly replace environment variables for the array yaml values. BUG, (thanks @lyt8384)

🧹 Chore:

  • 🧑‍🏭: [ALPHA] gRPC plugin: base64 decoder for the google's ErrorProto structure. FR, (thanks @rauanmayemir)

v2.11.2 (13.09.2022)

👀 New:

  • ✏ Kafka plugin: [ ⚠️ EXPERIMENTAL OPTION ⚠️ ] Kafka plugin now waits for 1 minute (automatically) for the broker to be available, FR, (thanks @Baiquette)
  • ✏ Internal: PHP Worker now uses an FSM to transition between states (working, ready, invalid, etc).
  • ✏ Internal: ./rr reset now works in parallel. All workers will be restarted simultaneously instead of a one-by-one sync approach.
  • ✏ Internal: ./rr reset and destroy (when stopping RR) now gracefully stop the workers (giving a chance for the finalizers to work). If the worker doesn't respond in 10 seconds, it'll be killed.

🩹 Fixes:

  • 🐛 SQS plugin: Incorrect detection of the AWS IMDSv2 instances, BUG (thanks @paulermo)
  • 🐛 Temporal plugin: Segmentation violation when using TLS, BUG, (thanks @seregazhuk)
  • 🐛 NATS plugin: Properly check the stream not found error from NATS, BUG, (thanks @pjtuxe)

🧹 Chore:

  • 🧑‍🏭: Temporal plugin: Support for the statsd daemon for stats aggregation, FR, (thanks @cv65kr) Configuration stays the same (no breaking changes), but additionally, you may specify a driver:

Prometheus:

temporal:
  address: "127.0.0.1:7233"
  metrics:
    driver: prometheus # <---- prometheus used by default (you may omit the driver in this case)
    address: "127.0.0.1:9095"
    prefix: "samples"
    type: "summary"
  activities:
    num_workers: 4

Statsd:

temporal:
  address: "127.0.0.1:7233"
  metrics:
    driver: statsd # <---- Should be specified to use a statsd driver
    host_port: "127.0.0.1:8125"
    prefix: "samples"
    flush_interval: 1s
    flush_bytes: 512
    tags:
      - foo: bar
  activities:
    num_workers: 4

Detailed description is here: link

v2.11.1 (25.08.2022)

👀 New:

  • ✏ http plugin: Send raw body (unescaped) to the PHP worker for the application/x-www-form-urlencoded content type. FR1, FR2, (thanks @ekisu, @rlantingmove4mobile) Configuration:
http:
  raw_body: true/false (by default)
  • ✏ temporal plugin: Overwrite client-name and client-version in Go client to represent PHP-SDK, FR, (thanks, @wolfy-j)

🧹 Chore:

  • 🧑‍🏭: Autocomplete .rr.yaml configuration for the cache plugin. link

v2.11.0 (18.08.2022)

⚠️ NewRelic middleware was removed. Please, use OTEL middleware instead

⚠️ In 2.12.0 we plan to replace websockets and broadcast plugins with the centrifuge plugin. However, if you still need a RR with these deprecated plugins, you may use Velox to build your custom build.

👀 New:

  • ✏️ [BETA]: RoadRunner: Can now be embedded in other go programs. PR, (thanks @khepin)
  • ✏️ gRPC Plugin: Implement Google's gRPC errors API. The exception might be passed as a Status structure in the Metadata (key - error) to be parsed and returned to the user w/o worker restart. NOTE: Status structure should be marshaled via proto marshaller, not json. FR
  • ✏️ Logger Plugin: Get rid of the context deadline exceeded error on worker's allocation. We updated the error message with the link to the docs with the most common causes for the worker allocation failed error: https://roadrunner.dev/docs/known-issues-allocate-timeout/2.x/en. Feel free to add your cases here :)
  • ✏️ CLI: New CLI command to pause, resume, destroy and list Jobs. FR, (thanks @hustlahusky)
  • ✏️ Velox: New configuration option: folder, which can be used to specify the folder with the plugin in the repository. (thanks, @darkweak)
  • ✏️ Velox: Velox now respects the plugin's replace directives. (thanks, @darkweak)
  • ✏️ Cache plugin: RR now uses a great cache (RFC 7234) plugin made by @darkweak
  • ✏️ [BETA] Kafka plugin: New Kafka driver for the Jobs plugin. FR, (thanks, @Smolevich)
  • ✏️ Temporal plugin: Temporal now uses a new reset mechanism to prevent WF worker restarts on activity worker failure
  • ✏️ Temporal plugin: Temporal plugin now supports a TSL-based authentication with the key and certificate.

Configuration:

temporal:
  tls:
    key: path/to/key
    cert: path/to/cert
  # other options

🩹 Fixes:

  • 🐛 Server plugin: use the allocate_timeout from the pool to wait for the tcp/unix socket connection from the PHP worker. BUG, (thanks @Warxcell)
  • 🐛 Velox: Fix panic when no github option is specified in the configuration.
  • 🐛 SDK: Use pool.allocate_timeout for the sockets/tcp relays instead of silently used of relay_timeout.

🧹 Chore:

  • 🧽 Logger plugin: use the parsable timestamp format for the raw logger mode. CH, (thanks @ilsenem)

🔧 Maintenance:

  • Temporal GO-SDK and API updated to the latest versions.
  • All plugins, including RR, now use Go 1.19

v2.10.7 (14.07.2022)

👀 New:

  • ✏️ OTEL Middleware: Support for the jaeger_agent exporter - BUG, (thanks @L3tum)

📦 Packages:

  • 📦 OTEL SDK updated to v1.8.0
  • 📦 google.golang.org/gRPC updated to v1.48.0
  • 📦 logger plugin updated to v2.13.5
  • 📦 Go updated to v1.18.4

v2.10.6 (07.07.2022)

🩹 Fixes:

  • 🐛 SDK: In some cases, worker watcher might freeze if the user kills the worker right after allocation but before wait4 syscall and become a zombie. BUG
  • 🐛 AMQP Plugin: Ignored prefetch option when dynamically creating a pipeline. BUG (thanks @rauanmayemir)

👀 New:

  • ✏️ VELOX: Velox now supports environment variables for the version, buildtime, GitHub, and GitLab tokens in the velox.toml.
  • ✏️ Logger: Say bye-bye to the CRC verification failed error. Starting from the v2.10.6, RR will show user-friendly message with the link to our docs on the most common causes for this type of error. Docs

v2.10.5 (23.06.2022)

🩹 Fixes:

  • 🐛 SDK: Increase stderr buffer size from 32kb to 65kb, man7, BUG, (thanks @7krasov)
  • 🐛 AMQP Plugin: Fix incorrect queue binding to the default routing key. BUG, (thanks @rauanmayemir)
  • 🐛 HTTP Plugin: Fix x-www-form-urlencoded requests 10Mb limit. BUG (thanks @StreetYo)

v2.10.4 (11.06.2022)

🩹 Fixes:

  • 🐛 Fix: incorrect reset behavior for the workers pool.
  • 🐛 Fix: correct reset order for the RR and Temporal workers.

v2.10.3 (02.06.2022)

👀 New:

  • ✏️ CLI: rr stop command. rr stop will read the .pid file to send a graceful stop signal to the main RR process (SIGTERM). FR (thanks @Baiquette)

🩹 Fixes:

  • 🐛 Fix: incorrect reset behavior for the temporal plugin. BUG, BUG, BUG. (thanks @dmitry-pilipenko, @mzavatsky)

v2.10.2 (26.05.2022)

👀 New:

  • ✏️ WORKER: Starting from this release, RR is able to show full error messages sent to the STDOUT during the worker bootstrap.

  • ✏️ HTTP: Connection might be upgraded from the http/1.1 to h2c: rfc7540 Headers, which should be sent to upgrade connection:

    1. Upgrade: h2c
    2. Connection: HTTP2-Settings
    3. Connection: Upgrade
    4. HTTP2-Settings: AAMAAABkAARAAAAAAAIAAAAA RFC

  • ✏️ VELOX: Add GitLab support. Starting from the beta.2 you may use GitHub and GitLab plugins together. Configuration updated. Keep in mind, until stable release 1.0.0 configuration might be changed with breaking changes.

  • ✏️ protoc-gen-php-grpc: Use of fully qualified names in place of imports. PR (thanks @ryanjcohen)

v2.10.1 (19.05.2022)

👀 New:

  • ✏️ Jobs (queues) plugin now can consume any payload from the queue. If RR fails in converting payload into the Job structure, it'll create and fill all available fields manually. To turn on this feature, use consume_all: true in the driver configuration, e.g.: Supported drivers: amqp, sqs,beanstalk,nats.
jobs:
  num_pollers: 10
  pipeline_size: 100000
  pool:
    num_workers: 10

  pipelines:
    test-raw:
      driver: sqs
      config:
        consume_all: true # <------- NEW OPTION

  consume: [ "test-raw" ]
  • ✏️ SQS Jobs driver now can skip queue declaration in favor of getting queue URL instead. To use this feature, use skip_queue_declaration: true sqs driver option. FR, (thanks @sergey-telpuk)
jobs:
  num_pollers: 10
  pipeline_size: 100000
  pool:
    num_workers: 10

  pipelines:
    test-2:
      driver: sqs
      config:
        skip_queue_declaration: true # <----- NEW OPTION

  consume: [ "test-2" ]
  • ✏️ OpenTelemetry middleware now supports Jaeger exporter and propagator.
http:
  address: 127.0.0.1:43239
  max_request_size: 1024
  middleware: [gzip, otel]
  pool:
    num_workers: 2
    max_jobs: 0
    allocate_timeout: 60s
    destroy_timeout: 60s

otel:
  exporter: jaeger # <----- NEW OPTION
  • ✏️ HTTP Plugin now supports mTLS authentication. Possible values for the client_auth_type are the same as for the gRPC (no_client_cert, request_client_cert,require_any_client_cert,verify_client_cert_if_given,require_and_verify_client_cert) FR, (thanks @fwolfsjaeger)
version: '2.7'

server:
  command: "php ../../php_test_files/http/client.php echo pipes"
  relay: "pipes"
  relay_timeout: "20s"

http:
  address: :8085
  max_request_size: 1024
  middleware: [ ]
  pool:
    num_workers: 1
    max_jobs: 0
    allocate_timeout: 60s
    destroy_timeout: 60s
  ssl:
    address: :8895
    key: "key"
    cert: "cert"
    root_ca: "rootCA.pem" # <---- REQUIRED to use mTLS
    client_auth_type: require_and_verify_client_cert # <---- NEW OPTION
logs:
  mode: development
  level: error

🩹 Fixes:

  • 🐛 Fix: HTTP plugin: non-documented behavior on non-standard (but valid) http codes. BUG, (thanks, @Meroje)
  • 🐛 Fix: SQS driver: rr_auto_ack attribute won't fail the existing messages.

v2.10.0 (16.05.2022)

👀 New:

  • ✏️ Documentation update: link.
  • ✏️ RoadRunner-Temporal plugin now supports local activities. Here is the brief overview: link.
  • ✏️ Add Debian amd64 releases. FEATURE
  • ✏️ Add signed releases. Starting from the v2.10.0, every released binary can be checked with a provided *.asc key. For example:
$ gpg --verify rr.asc

The openPGP key can be verified here: keyserver

  • ✏️ All proto api for the Go programming language located here: link. To use it, just import the latest stable version go.buf.build/protocolbuffers/go/roadrunner-server/api latest.
  • ✏️ Service plugin now supports auto-reload. It can be added to the reload plugin targets and on change, it'll reload all underlying processes.
  • ✏️ AutoAck jobs option. For the messages (jobs), which are acceptable to lose. Or which execution can lead to a worker's stop (for example - OOM).
  • ✏️ [BETA] OpenTelemetry support. Starting from now, the new_relic middleware is deprecated, it'll receive only dependency updates and will be removed from the RR bundle in the v2.12.0. (thanks @brettmc) OpenTelemetry plugin supports the following exporters:
    1. OTLP (open telemetry protocol): datadog, new relic.
    2. zipkin
    3. stdout All these exporters can send their data via http or grpc clients.

Configuration sample (stdout exporter):

http:
  address: 127.0.0.1:43239
  max_request_size: 1024
  middleware: [gzip, otel]
  pool:
    num_workers: 2
    max_jobs: 0
    allocate_timeout: 60s
    destroy_timeout: 60s

otel:
  insecure: false
  compress: true
  exporter: stdout
  service_name: rr_test
  service_version: 1.0.0

New Relic exporter via http client: link

http:
    address: 127.0.0.1:43239
    max_request_size: 1024
    middleware: [gzip, otel]
    pool:
        num_workers: 2
        max_jobs: 0
        allocate_timeout: 60s
        destroy_timeout: 60s
otel:
  insecure: false
  compress: true
  client: http
  exporter: stdout
  custom_url: ""
  service_name: rr_test
  service_version: 1.0.0
  endpoint: otlp.eu01.nr-data.net:4318
  headers:
    - api-key: xxx # your api key here

PHP worker can access tracing data via w3c headers.

  • ✏️ Jobs can be auto-ack'ed now. New option used to acknowledge a message after RR receive it from the queue. FR, (thanks @hustlahusky)
  • ✏️ protoc-gen-php-grpc now supports optional fields. (thanks @genhoi)

🧹 Chore:

  • 🧑‍🏭: All spaces and new-lines from the Service plugin output will be automatically trimmed. CHORE, (thanks, @OO00O0O)

v2.9.4 (06.05.2022)

🩹 Fixes:

  • 🐛 Fix: HTTP plugin: request max body size incorrectly parsed. BUG, (thanks, @Opekunov)

v2.9.3 (06.05.2022)

👀 New:

  • ✏️: --silent flag. This flag will hide startup message and ./rr reset output.

🩹 Fixes:

  • 🐛 Fix: AMQP driver didn't reconnect on timeouts, which led to stopping consuming messages w/o a proper notification. BUG, (thanks @hustlahusky)
  • 🐛 Fix: reset command (./rr reset) gets stuck when using output redirects. BUG, (thanks @maximal)

🧹 Chore:

  • 🏗️ HTTP plugin: better looking error message on wrong status code. ISSUE, (thanks @gam6itko)

v2.10.0-alpha.1 (07.04.2022)

👀 New:

  • ✏️ [ALPHA] HTTP response streaming. Available only in the alfa builds.

Worker sample:

<?php

use Nyholm\Psr7\Factory\Psr17Factory;
use Nyholm\Psr7\Response;
use Nyholm\Psr7\Stream;
use Spiral\RoadRunner;

ini_set('display_errors', 'stderr');
require __DIR__ . "/vendor/autoload.php";

$worker = RoadRunner\Worker::create();
$psr7 = new RoadRunner\Http\PSR7Worker(
    $worker,
    new Psr17Factory(),
    new Psr17Factory(),
    new Psr17Factory()
);

$psr7->chunk_size = 10 * 10 * 1024;
$filename = 'file.tmp'; // big file or response

while ($req = $psr7->waitRequest()) {
    try {
        $fp = \fopen($filename, 'rb');
        \flock($fp, LOCK_SH);
        $resp = (new Response())->withBody(Stream::create($fp));
        $psr7->respond($resp);
    } catch (\Throwable $e) {
        $psr7->getWorker()->error((string)$e);
    }

Known issues:

  1. RR will not notify a worker if HTTP connection was interrupted. RR will read all response from the worker and drop it. That will be fixed in the stable streaming release.
  2. Sometimes RR may miss the immediate error from the worker and send a 0 payload with 200 status. This is related only to the http response.
  • ✏️ [BETA] Local activities support link.

v2.9.2 (28.04.2022)

🩹 Fixes:

  • 🐛 Fix: incorrect -w flag behavior (set working dir). BUG, (thanks @rauanmayemir)

v2.9.1 (11.04.2022)

⚠️ This is important. trusted_subnets HTTP option will work only when proxy_ip_parser was added to the list of the middlewares.

🩹 Fixes:

  • 🐛 Fix: incorrect usage of the ExecWithTTL SDK method which leads to skipping the exec_ttl supervisor option BUG (thanks @Baiquette)

👀 New:

  • ✏️ Add proxy_ip_parser HTTP middleware responsible for parsing X-Forwarded-For, Forwarded, True-Client-IP and X-Real-Ip.

v2.9.0 (07.04.2022)

🇺🇦🇺🇦🇺🇦 #StandWithUkraine 🇺🇦🇺🇦🇺🇦

👀 New:

  • ✏️ API: add service proto api to manage services, FR (thanks @butschster). Documentation is here: link.
  • ✏️ Grafana dashboard PATH. Exposed metrics:
  1. General:
    1. Uptime (seconds).
    2. Memory used by RR (MB).
    3. Number of active goroutines.
  2. HTTP:
    1. Number of workers by its state (ready, working, invalid).
    2. Total RSS memory used by workers.
    3. Memory used by each worker individually (with PID).
    4. Latency (ms).
    5. Requests queue size.
    6. Requests per minute.
  3. JOBS:
    1. Number of workers by its state (ready, working, invalid).
    2. Total RSS memory used by workers.
    3. Memory used by each worker individually (with PID).
    4. Successfully processed jobs (rate, 5m).
    5. Failed jobs (rate, 5m).

🩹 Fixes:

  • 🐛 Fix: Goroutines leak in the amqp plugin when destroying pipelines.

🧹 Chore:

  • 🧑‍🏭 Update all dependencies to the most recent versions.

v2.8.8 (31.03.2022)

🩹 Fixes:

  • 🐛 Fix: Chdir (if the user sets the -w flag) before searching for the configuration.

📦 Packages:

  • 📦 Update SQS and Go deps to the most recent versions.

v2.8.6 (24.03.2022)

🩹 Fixes:

  • 🐛 Fix: Websocket access validator requests don't send cookies BUG, (@steffendietz)

v2.8.5 (23.03.2022)

🧹 Chore:

  • 🧑‍🏭 Update all dependencies to the most recent versions.

🩹 Fixes:

  • 🐛 Fix: Incorrect pointer assign to read the configuration BUG

v2.8.4 (17.03.2022)

👀 New:

  • Go update to version 1.18

🧹 Chore:

  • 🧑‍🏭 Update all dependencies to the most recent versions.

🩹 Fixes:

  • 🐛 Fix: No longer able to set environment variables for service BUG, (reporter @andrei-dascalu)

v2.8.3 (13.03.2022)

👀 New:

  • ✏️ Better env variables parser. Now RR is able to parse the sentences like: "mysql://${MYSQL_USER}:${MYSQL_PASSWORD}@${MYSQL_HOST}:${MYSQL_PORT}/${MYSQL_DATABASE}?serverVersion=5.7" and get all environment variables. FR, (reporter @Tony-Sol)

🧹 Chore:

  • 🧑‍🏭 Update all dependencies to the most recent versions.
  • 🧑‍🏭 Remove configuration plugin from the root.go and other files. Used only in the serve where it should be.

🩹 Fixes:

  • 🐛 Fix: call of the kv.TTL for the Redis drivers returns non RFC3339 time format BUG, (reporter @antikirra)
  • 🐛 Fix: rr workers command doesn't work for the service plugin BUG, (reporter @OO00O0O)

v2.8.2 (22.02.2022)

🧹 Chore:

  • Docker: migrate to the trivy action instead of grype PR, (contributor: @tarampampam)

🩹 Fixes:

  • 🐛 Fix: CONFIG: version can't be passed as inline option: BUG, (reporter @nunomaduro)

v2.8.1 (21.02.2022)

🩹 Fixes:

  • 🐛 Fix: HTTP: incorrect middleware chain order BUG, (reporter @nitrogenium)
http:
  middleware: ["static", "gzip"]

Middleware applied from the right to left, i.e. first will be gzip and then static.

👀 New:

  • ✏️ HTTP: Properly parse Forwarder header, RFC, FR (reporter @digitalkaoz)
  • ✏️ TEMPORAL: Add rr_activities_pool_queue_size and rr_workflows_pool_queue_size metrics. These metrics shows the number of activities/workflows waiting for the worker BUG, (reporter @Zylius)
  • ✏️ API: Queuer interface for the workers pool to show the number of requests waiting for the worker. Implemented in SDK.

v2.8.0 (17.02.2022)

⚠️ For the RR v2.8.0 your .rr.yaml configuration version should be 2.7, so please update your config and add version: '2.7' at the top.

👀 New:

  • ✏️ Timeout for the SQS create/get queue operations (30seconds) FR (reporter @sergey-telpuk)
  • ✏️ New workers pool option -> command:
http:
    address: 127.0.0.1:8080
    max_request_size: 256
    middleware: ["headers", "gzip"]
    trusted_subnets: []

    # Workers pool settings.
    pool:
        command: "php app.php" <---- NEW

This option is used to override the server's command with the new one. The new command will inherit all server options. It can be used in any plugin (jobs, grpc, tcp, http, etc) using workers pool (http used here as a sample). No need to update your config. By default server command will be used as in the previous RR versions. Note, you can't leave the server command option empty. If you wanted to override the command in every plugin you use, put a placeholder in the server.command.

  • ✏️ Add Rr_newrelic_ignore header support. Now you can send the Rr_newrelic_ignore:true header and RR will ignore such transactions. No data will be sent to the newrelic server. FR (reporter @arku31)
  • ✏️ Add support for the SQS FIFO queues FR (reporter @paulermo)
  • ✏️ Add support for the gRPC Healthcheck protocol v1 in the grpc plugin. Doc, FR (reporter @porozhnyy)

v2.7.9 (14.02.2022)

🩹 Fixes:

  • 🐛 Fix: errors on workers reallocating when the unix/tcp socket transport is used: BUG, (reporter @tarampampam)

2.7.8 (12.02.2022)

🩹 Fixes:

  • 🐛 Fix: worker sometimes panics when unix/tcp sockets transport used: BUG, (reporter @tarampampam)

2.7.7 (10.02.2022)

🩹 Fixes:

  • 🐛 Fix: case-sensitive attributes for SQS BUG, (reporter @paulermo)
  • 🐛 Fix: grpc plugin incorrectly parses the error response from the worker BUG, (reporter @rauanmayemir)
  • 🐛 Fix: destroy operation hangs if the number of workers is 0 BUG, (reporter @benalf)

2.7.6 (06.02.2022)

🩹 Fixes:

  • 🐛 Fix: temporal: ActivityNotRegisteredError on the local activity workers BUG (reporter @Zylius)

2.7.5 (04.02.2022)

👀 New:

  • ✏️ Better workers' statistic. Add worker state, per-worker memory usage and worker's PID stats for the JOBS, GRPC, HTTP plugins FR

2.7.4 (27.01.2022)

👀 New:

  • ✏️ Return an error if the user uses the .env file, but it doesn't exist. BUG, (reporter @O00O0O)
  • ✏️ Parallel workers allocation within the plugin. FR, (reporter @roquie)

🩹 Fixes:

  • 🐛 Fix: RR workers are blocked when one of them should send a big (> 100mb) response and when the time to send the response is bigger than the supervisor's watch_tick BUG (reported by: @OO00O0O)

2.7.3 (19.01.2022)

🩹 Fixes:

  • 🐛 Fix: jobs pipeline shadowing if registering the pipeline with the existing name BUG
  • 🐛 Fix: reload plugin doesn't collect a resettable plugins BUG

v2.7.2 (17.01.2022)

👀 New:

  • Moved to the new organization.

v2.7.0 (14.01.2022)

👀 New:

  • ✏️ RR workers pool, worker, worker_watcher now has their own log levels. stderr/stdout logged as before at the info log level. All other messages moved to the debug log level except a few events from the worker_watcher when RR can't allocate the new worker which are moved to the warn.
  • ✏️ Use the common logger for the whole roadrunner-sdk and roadrunner-plugins.
  • ✏️ .rr.yaml now support versions. You may safely use your old configurations w/o specifying versions. Configuration w/o version will be treated as 2.6. It is safe to use configuration w/o version or with version 2.6 with RR 2.7 because RR is able to automatically transform the old configuration. But if you use configuration version 2.7 you must update the jobs pipelines config. At this point we can guarantee, that no breaking changes will be introduced in the configuration w/o auto-convert from the older configuration version For example, if we introduce a configuration update let's say in version 2.10, we will support automatic conversion from at least 2 previous versions w/o involving the user into the process. In the example case, versions 2.9 and 2.8 will be automatically converted. From our release cycle, you will have at least 3 months to update the configuration from version 2.8 and 2 months from 2.9.Version located at the top of the .rr.yaml:

Compatibility matrix located here: TODO Configuration changelog: TODO

version: "2.6"

# ..... PLUGINS ......

Before:

  pipelines:
    test-local:
      driver: memory
      priority: 10
      prefetch: 10000

    test-local-1:
      driver: boltdb
      priority: 10
      file: "rr.db"
      prefetch: 10000

    test-local-2:
      driver: amqp
      prefetch: 10
      priority: 1
      queue: test-1-queue
      exchange: default
      exchange_type: direct
      routing_key: test
      exclusive: false
      multiple_ack: false
      requeue_on_fail: false

    test-local-3:
      driver: beanstalk
      priority: 11
      tube_priority: 1
      tube: default-1
      reserve_timeout: 10s

    test-local-4:
      driver: sqs
      priority: 10
      prefetch: 10
      visibility_timeout: 0
      wait_time_seconds: 0
      queue: default
      attributes:
        DelaySeconds: 0
        MaximumMessageSize: 262144
        MessageRetentionPeriod: 345600
        ReceiveMessageWaitTimeSeconds: 0
        VisibilityTimeout: 30
      tags:
        test: "tag"

    test-local-5:
      driver: nats
      priority: 2
      prefetch: 100
      subject: default
      stream: foo
      deliver_new: true
      rate_limit: 100
      delete_stream_on_stop: false
      delete_after_ack: false

After: Now, pipelines have only driver key with the configuration under the config key. We did that to uniform configuration across all drivers (like in the KV).

  pipelines:
    test-local:
      driver: memory

      config: # <------------------ NEW
        priority: 10
        prefetch: 10000

    test-local-1:
      driver: boltdb

      config: # <------------------ NEW
        priority: 10
        file: "test-local-1-bolt.db"
        prefetch: 10000

    test-local-2:
      driver: amqp

      config: # <------------------ NEW
        priority: 11
        prefetch: 100
        queue: test-12-queue
        exchange: default
        exchange_type: direct
        routing_key: test
        exclusive: false
        multiple_ack: false
        requeue_on_fail: false

    test-local-3:
      driver: beanstalk

      config: # <------------------ NEW
        priority: 11
        tube_priority: 1
        tube: default-2
        reserve_timeout: 10s

    test-local-4:
      driver: sqs

      config: # <------------------ NEW
        priority: 10
        prefetch: 10
        visibility_timeout: 0
        wait_time_seconds: 0
        queue: default

        attributes:
          DelaySeconds: 0
          MaximumMessageSize: 262144
          MessageRetentionPeriod: 345600
          ReceiveMessageWaitTimeSeconds: 0
          VisibilityTimeout: 30
        tags:
        test: "tag"

    test-local-5:
      driver: nats

      config: # <------------------ NEW
        priority: 2
        prefetch: 100
        subject: default
        stream: foo
        deliver_new: true
        rate_limit: 100
        delete_stream_on_stop: false
        delete_after_ack: false
  • ✏️ [ALPHA] New cache http middleware. It is still in alpha, but we started implementing the rfc-7234 to support Cache-Control and caching in general. In the first alpha you may test the max-age, Age and Authorization support via the in-memory driver.

Configuration:

http:
# .....
    middleware: ["cache"]
    cache:
        driver: memory
        cache_methods: ["GET", "HEAD", "POST"] # only GET in alpha
        config: {} # empty configuration for the memory
  • ✏️ Logger unification. Starting this version we bound our logs to the uber/zap log library as one of the most popular and extensible.
  • ✏️ API stabilization. All v2 api interfaces moved to the https://github.com/roadrunner-server/api repository. Except logger (structure), all plugins depends only on the interfaces and don't import each other.
  • ✏️ GRPC plugin now is able to work with gzipped payloads. FR (reporter @hetao29)
  • ✏️ SQS plugin now detects EC2 env and uses AWS credentials instead of static provider. FR (reporter @paulermo)
  • ✏️ Jobs plugin now acknowledges responses with incorrectly formed responses to prevent the infinity loop (with the error message in the logs). BUG (reporter @sergey-telpuk)
  • ✏️ protoc updated to the version v3.19.2.

🩹 Fixes:

  • 🐛 Fix: RR may have missed the message from the stderr when the PHP script failed to start immediately after RR starts.
  • 🐛 Fix: 200 HTTP status code instead of 400 on readiness/health bad requests. BUG
  • 🐛 Fix: new_relic plugin removes/modifies more headers than it should. BUG (reporter: @arku31)

v2.6.6 (7.12.2021)

👀 New:

  • ✏️ Add events from the supervisor to the server plugin.

🩹 Fixes:

  • 🐛 Fix: worker exited immediately after obtaining the response. BUG (reporter: @samdark).

📦 Packages:

  • 📦 Update RoadRunner to v2.6.2

v2.6.5 (7.12.2021)

🩹 Fixes:

  • 🐛 Fix: wrong metrics type for the rr_http_requests_queue, bug (reporter: @victor-sudakov)
  • 🐛 Fix: memory leak when supervised static pool used. PR.

📦 Packages:

  • 📦 Update RoadRunner to v2.6.1

v2.6.4 (7.12.2021)

📦 Packages:

  • 📦 Update endure to v1.1.0

🩹 Fixes:

  • 🐛 Fix: NPE in the http.Reset. BUG

v2.6.3 (3.12.2021)

👀 New:

  • ✏️ informer.List RPC call return all available plugins with workers instead of all available plugins. This behavior was changed because Informer has the dependency of every RR plugin, which led to the cycles. This is not an external API and used only internally.
  • ✏️ Beanstalk queue returned to the [ALPHA] stage. It's very unstable when destroying pipelines and can lead to infinite read loops when something wrong with the connection. Use with care.
  • ✏️ Go version updated to v1.17.4.

🩹 Fixes:

  • 🐛 Fix: add missing plugins to the container: fileserver, http_metrics.

v2.6.2 (3.12.2021)

🩹 Fixes:

  • 🐛 Fix: Random NPE on RR start. BUG

v2.6.1 (2.12.2021)

🩹 Fixes:

  • 🐛 Fix: logger incorrectly escaped HTML, JSON, and other special symbols.

v2.6.0 (30.11.2021)

👀 New:

  • ✏️ New internal message bus. Available globally. Supports wildcard subscriptions (for example: http.* will subscribe you to the all events coming from the http plugin). The subscriptions can be made from any RR plugin to any RR plugin.
  • ✏️ Now, RR will show in the returned error the bad header content in case of CRC mismatch error. More info in the PR.
  • ✏️ [BETA] Support for the New Relic observability platform. Sample of the client library might be found here. (Thanks @arku31) New Relic middleware is a part of the HTTP plugin, thus configuration should be inside it:
http:
  address: 127.0.0.1:15389
  middleware: [ "new_relic" ] <------- NEW
  new_relic: <---------- NEW
    app_name: "app"
    license_key: "key"
  pool:
    num_workers: 10
    allocate_timeout: 60s
    destroy_timeout: 60s

License key and application name could be set via environment variables: (leave app_name and license_key empty)

  • license_key: NEW_RELIC_LICENSE_KEY.
  • app_name: NEW_RELIC_APP_NAME.

To set the New Relic attributes, the PHP worker should send headers values withing the rr_newrelic header key. Attributes should be separated by the :, for example foo:bar, where foo is a key and bar is a value. New Relic attributes sent from the worker will not appear in the HTTP response, they will be sent directly to the New Relic.

To see the sample of the PHP library, see the @arku31 implementation: https://github.com/arku31/roadrunner-newrelic

The special key which PHP may set to overwrite the transaction name is: transaction_name. For example: transaction_name:foo means: set transaction name as foo. By default, RequestURI is used as the transaction name.

        $resp = new \Nyholm\Psr7\Response();
        $rrNewRelic = [
            'shopId:1', //custom data
            'auth:password', //custom data
            'transaction_name:test_transaction' //name - special key to override the name. By default it will use requestUri.
        ];

        $resp = $resp->withHeader('rr_newrelic', $rrNewRelic);
  • ✏️ [BETA] New plugin: TCP. The TCP plugin is used to handle raw TCP payload with a bi-directional protocol between the RR server and PHP worker.

PHP client library: https://github.com/spiral/roadrunner-tcp

Configuration:

rpc:
  listen: tcp://127.0.0.1:6001

server:
  command: "php ../../psr-worker-tcp-cont.php"

tcp:
  servers:
    server1:
      addr: 127.0.0.1:7778
      delimiter: "\r\n"
    server2:
      addr: 127.0.0.1:8811
      read_buf_size: 10
    server3:
      addr: 127.0.0.1:8812
      delimiter: "\r\n"
      read_buf_size: 1

  pool:
    num_workers: 5
    max_jobs: 0
    allocate_timeout: 60s
    destroy_timeout: 60s
  • ✏️ New HTTP middleware: http_metrics.
http:
  address: 127.0.0.1:15389
  middleware: [ "http_metrics" ] <------- NEW
  pool:
    num_workers: 10
    allocate_timeout: 60s
    destroy_timeout: 60s

All old and new http metrics will be available after the middleware is activated. Be careful, this middleware may slow down your requests. New metrics:

- `rr_http_requests_queue_sum` - number of queued requests.
- `rr_http_no_free_workers_total` - number of the occurrences of the `NoFreeWorkers` errors.
  • ✏️ New file server to serve static files. It works on a different address, so it doesn't affect the HTTP performance. It uses advanced configuration specific for the static file servers. It can handle any number of directories with its own HTTP prefixes. Config:
fileserver:
  # File server address
  #
  # Error on empty
  address: 127.0.0.1:10101
  # Etag calculation. Request body CRC32.
  #
  # Default: false
  calculate_etag: true

  # Weak etag calculation
  #
  # Default: false
  weak: false

  # Enable body streaming for the files more than 4KB
  #
  # Default: false
  stream_request_body: true

  serve:
    # HTTP prefix
    #
    # Error on empty
  - prefix: "/foo"

    # Directory to serve
    #
    # Default: "."
    root: "../../../tests"

    # When set to true, the server tries minimizing CPU usage by caching compressed files
    #
    # Default: false
    compress: false

    # Expiration duration for inactive file handlers. Units: seconds.
    #
    # Default: 10, use a negative value to disable it.
    cache_duration: 10

    # The value for the Cache-Control HTTP-header. Units: seconds
    #
    # Default: 10 seconds
    max_age: 10

    # Enable range requests
    # https://developer.mozilla.org/en-US/docs/Web/HTTP/Range_requests
    #
    # Default: false
    bytes_range: true

  - prefix: "/foo/bar"
    root: "../../../tests"
    compress: false
    cache_duration: 10s
    max_age: 10
    bytes_range: true
  • ✏️ on_init option for the server plugin. on_init code executed before the regular command and can be used to warm up the application for example. Failed on_init command doesn't affect the main command, so, the RR will continue to run. Thanks (@OO00O0O)

Config:

# Application server settings (docs: https://roadrunner.dev/docs/php-worker)
server:
  on_init: <----------- NEW
    # Command to execute before the main server's command
    #
    # This option is required if using on_init
    command: "any php or script here"

    # Script execute timeout
    #
    # Default: 60s [60m, 60h], if used w/o units its means - NANOSECONDS.
    exec_timeout: 20s

    # Environment variables for the worker processes.
    #
    # Default: <empty map>
    env:
      - SOME_KEY: "SOME_VALUE"
      - SOME_KEY2: "SOME_VALUE2"

  # ..REGULAR SERVER OPTIONS...
  • ✏️ [BETA] GRPC can handle multiply proto files. Config:
# GRPC service configuration
grpc:
    # Proto files to use
    #
    # This option is required. At least one proto file must be specified.
    proto:
        - "first.proto"
        - "second.proto"

## ... OTHER REGULAR GRPC OPTIONS ...
  • ✏️ New allow configuration option for the http.uploads and multipart requests. The new option allows you to filter upload extensions knowing only allowed. Now, there is no need to have a looong list with all possible extensions to forbid. FR (Thanks @rjd22) http.uploads.forbid has a higher priority, so, if you have duplicates in the http.uploads.allow and http.uploads.forbid the duplicated extension will be forbidden. Config:
http:
  address: 127.0.0.1:18903
  max_request_size: 1024
  middleware: ["pluginMiddleware", "pluginMiddleware2"]
  uploads:
    forbid: [".php", ".exe", ".bat"]
    allow: [".html", ".aaa" ] <------------- NEW
  trusted_subnets:
    [
      "10.0.0.0/8",
      "127.0.0.0/8",
      "172.16.0.0/12",
      "192.168.0.0/16",
      "::1/128",
      "fc00::/7",
      "fe80::/10",
    ]
  pool:
    num_workers: 2
    max_jobs: 0
    allocate_timeout: 60s
    destroy_timeout: 60s

  • ✏️ Beanstalk queue reject stop RPC calls if there are jobs in the priority queue associated with the requested pipeline.

  • ✏️ Startup message when the RR has started.

🩹 Fixes:

  • 🐛 Fix: GRPC server will show message when started.
  • 🐛 Fix: Static plugin headers were added to all requests. BUG
  • 🐛 Fix: zombie processes in the pool.debug mode.

📦 Packages:

  • 📦 roadrunner v2.6.0
  • 📦 roadrunner-plugins v2.6.0
  • 📦 roadrunner-temporal v1.0.11
  • 📦 endure v1.0.8
  • 📦 goridge v3.2.4
  • 📦 temporal.io/sdk v1.11.1

v2.5.3 (27.10.2021)

🩹 Fixes:

  • 🐛 Fix: panic in the TLS layer. The http plugin used http server instead of https in the rootCA routine.

v2.5.2 (23.10.2021)

🩹 Fixes:

  • 🐛 Fix: ASLR builds causes problems in the docker.

v2.5.1 (22.10.2021)

🩹 Fixes:

  • 🐛 Fix: base64 response instead of json in some edge cases.

v2.5.0 (20.10.2021)

💔 Breaking change:

  • 🔨 Some drivers now use a new config key to handle local configuration. Involved plugins and drivers:
  • plugins: broadcast, kv
  • drivers: memory, redis, memcached, boltdb.

Old style:

broadcast:
    default:
        driver: memory
        interval: 1

New style:

broadcast:
  default:
    driver: memory
     config: {} <--------------- NEW
kv:
  memory-rr:
    driver: memory
    config: <--------------- NEW
      interval: 1

kv:
  memcached-rr:
    driver: memcached
    config: <--------------- NEW
       addr:
         - "127.0.0.1:11211"

broadcast:
  default:
    driver: redis
    config: <------------------ NEW
      addrs:
        - "127.0.0.1:6379"

👀 New:

  • ✏️ [BETA] GRPC plugin update to v2.
  • ✏️ Roadrunner-plugins repository. This is the new home for the roadrunner plugins with documentation, configuration samples, and common problems.
  • ✏️ [BETA] Let's Encrypt support. RR now can obtain an SSL certificate/PK for your domain automatically. Here is the new configuration:
    ssl:
      # Host and port to listen on (eg.: `127.0.0.1:443`).
      #
      # Default: ":443"
      address: "127.0.0.1:443"

      # Use ACME certificates provider (Let's encrypt)
      acme:
        # Directory to use as a certificate/pk, account info storage
        #
        # Optional. Default: rr_cache
        certs_dir: rr_le_certs

        # User email
        #
        # Used to create LE account. Mandatory. Error on empty.
        email: you-email-here@email

        # Alternate port for the http challenge. Challenge traffic should be redirected to this port if overridden.
        #
        # Optional. Default: 80
        alt_http_port: 80,


        # Alternate port for the tls-alpn-01 challenge. Challenge traffic should be redirected to this port if overridden.
        #
        # Optional. Default: 443.
        alt_tlsalpn_port: 443,

        # Challenge types
        #
        # Optional. Default: http-01. Possible values: http-01, tlsalpn-01
        challenge_type: http-01

        # Use production or staging endpoints. NOTE, try to use the staging endpoint (`use_production_endpoint`: `false`) to make sure, that everything works correctly.
        #
        # Optional, but for production should be set to true. Default: false
        use_production_endpoint: true

        # List of your domains to obtain certificates
        #
        # Mandatory. Error on empty.
        domains: [
            "your-cool-domain.here",
            "your-second-domain.here"
        ]
  • ✏️ Add a new option to the logs plugin to configure the line ending. By default, used \n.

New option:

# Logs plugin settings
logs:
    (....)
    # Line ending
    #
    # Default: "\n".
    line_ending: "\n"
http:
  address: 127.0.0.1:55555
  max_request_size: 1024
  access_logs: true <-------- Access Logs ON/OFF
  middleware: []

  pool:
    num_workers: 2
    max_jobs: 0
    allocate_timeout: 60s
    destroy_timeout: 60s
  • ✏️ HTTP middleware to handle X-Sendfile header. Middleware reads the file in 10MB chunks. So, for example for the 5Gb file, only 10MB of RSS will be used. If the file size is smaller than 10MB, the middleware fits the buffer to the file size.
http:
  address: 127.0.0.1:44444
  max_request_size: 1024
  middleware: ["sendfile"] <----- NEW MIDDLEWARE

  pool:
    num_workers: 2
    max_jobs: 0
    allocate_timeout: 60s
    destroy_timeout: 60s
  • ✏️ Service plugin now supports env variables passing to the script/executable/binary/any like in the server plugin:
service:
  some_service_1:
    command: "php test_files/loop_env.php"
    process_num: 1
    exec_timeout: 5s # s,m,h (seconds, minutes, hours)
    remain_after_exit: true
    env:  <----------------- NEW
      foo: "BAR"
    restart_sec: 1
  • ✏️ Server plugin can accept scripts (sh, bash, etc) in it's command configuration key:
server:
    command: "./script.sh OR sh script.sh" <--- UPDATED
    relay: "pipes"
    relay_timeout: "20s"

The script should start a worker as the last command. For the pipes, scripts should not contain programs, which can close stdin, stdout or stderr.

  • ✏️ Nats jobs driver support - PR.
nats:
  addr: "demo.nats.io"

jobs:
  num_pollers: 10
  pipeline_size: 100000
  pool:
    num_workers: 10
    max_jobs: 0
    allocate_timeout: 60s
    destroy_timeout: 60s

  pipelines:
    test-1:
      driver: nats
      prefetch: 100
      subject: "default"
      stream: "foo"
      deliver_new: "true"
      rate_limit: 100
      delete_stream_on_stop: false
      delete_after_ack: false
      priority: 2

  consume: [ "test-1" ]

  • Driver uses NATS JetStream API and is not compatible with non-js API.

  • ✏️ Response API for the NATS, RabbitMQ, SQS and Beanstalk drivers. This means, that you'll be able to respond to a specified in the response queue. Limitations:

    • To send a response to the queue maintained by the RR, you should send it as a Job type. There are no limitations for the responses into the other queues (tubes, subjects).
    • Driver uses the same endpoint (address) to send the response as specified in the configuration.

🩹 Fixes:

  • 🐛 Fix: local and global configuration parsing.
  • 🐛 Fix: boltdb-jobs connection left open after RPC close command.
  • 🐛 Fix: close beanstalk connection and release associated resources after pipeline stopped.
  • 🐛 Fix: grpc plugin fails to handle requests after calling reset.
  • 🐛 Fix: superfluous response.WriteHeader call when connection is broken.

📦 Packages:

  • 📦 roadrunner v2.5.0
  • 📦 roadrunner-plugins v2.5.0
  • 📦 roadrunner-temporal v1.0.10
  • 📦 endure v1.0.6
  • 📦 goridge v3.2.3

v2.4.1 (13.09.2021)

🩹 Fixes:

  • 🐛 Fix: bug with not-idempotent call to the attributes.Init.
  • 🐛 Fix: memory jobs driver behavior. Now memory driver starts consuming automatically if the user consumes the pipeline in the configuration.

v2.4.0 (02.09.2021)

💔 Internal BC:

  • 🔨 Pool, worker interfaces: payload now passed and returned by the pointer.

👀 New:

  • ✏️ Long-awaited, reworked Jobs plugin with pluggable drivers. Now you can allocate/destroy pipelines in the runtime. Drivers included in the initial release: RabbitMQ (0-9-1), SQS v2, beanstalk, memory and local queue powered by the boltdb. PR
  • ✏️ Support for the IPv6 (tcp|http(s)|empty [::]:port, tcp|http(s)|empty [::1]:port, tcp|http(s)|empty :// [0:0:0:0:0:0:0:1]:port) for RPC, HTTP and other plugins. RFC
  • ✏️ Support for the Docker images via GitHub packages.
  • ✏️ Go 1.17 support for the all spiral packages.

🩹 Fixes:

  • 🐛 Fix: fixed bug with goroutines waiting on the internal worker's container channel, issue.
  • 🐛 Fix: RR become unresponsive when new workers failed to re-allocate, issue.
  • 🐛 Fix: add debug pool config key to the .rr.yaml configuration reference.

📦 Packages:

  • 📦 Update goridge to v3.2.1
  • 📦 Update temporal to v1.0.9
  • 📦 Update endure to v1.0.4

📈 Summary:

v2.3.2 (14.07.2021)

🩹 Fixes:

  • 🐛 Fix: Do not call the container's Stop method after the container stopped by an error.
  • 🐛 Fix: Bug with ttl incorrectly handled by the worker PR
  • 🐛 Fix: Add RR_BROADCAST_PATH to the websockets plugin PR

📈 Summary:

v2.3.1 (30.06.2021)

👀 New:

  • ✏️ Rework broadcast plugin. Add architecture diagrams to the doc folder. PR
  • ✏️ Add Clear method to the KV plugin RPC. PR

🩹 Fixes:

  • 🐛 Fix: Bug with channel deadlock when exec_ttl was used and TTL limit reached PR
  • 🐛 Fix: Bug with healthcheck endpoint when workers were marked as invalid and stay is that state until next request PR
  • 🐛 Fix: Bugs with boltdb storage: Boom , Boom, Boom
  • 🐛 Fix: Bug with incorrect redis initialization and usage Bug
  • 🐛 Fix: Bug, Goridge duplicate error messages Bug
  • 🐛 Fix: Bug, incorrect request origin check Bug

📦 Packages:

  • 📦 Update goridge to v3.1.4
  • 📦 Update temporal to v1.0.8

📈 Summary:

v2.3.0 (08.06.2021)

👀 New:

  • ✏️ Brand new broadcast plugin now has the name - websockets with broadcast capabilities. It can handle hundreds of thousands websocket connections very efficiently (~300k messages per second with 1k connected clients, in-memory bus on 2CPU cores and 1GB of RAM) Issue
  • ✏️ Protobuf binary messages for the websockets and kv RPC calls under the hood. Issue
  • ✏️ Json-schemas for the config file v1.0 (it also registered in schemastore.org)
  • ✏️ latest docker image tag supported now (but we strongly recommend using a versioned tag (like 0.2.3) instead)
  • ✏️ Add new option to the http config section: internal_error_code to override default (500) internal error code. Issue
  • ✏️ Expose HTTP plugin metrics (workers memory, requests count, requests duration) . Issue
  • ✏️ Scan server.command and find errors related to the wrong path to a PHP file, or .ph, .sh scripts. Issue
  • ✏️ Support file logger with log rotation Wiki , Issue

🩹 Fixes:

  • 🐛 Fix: Bug with informer.Workers worked incorrectly: Bug
  • 🐛 Fix: Internal error messages will not be shown to the user (except HTTP status code). Error message will be in logs: Bug
  • 🐛 Fix: Error message will be properly shown in the log in case of SoftJob error: Bug
  • 🐛 Fix: Wrong applied middlewares for the fcgi server leads to the NPE: Bug

📦 Packages:

  • 📦 Update goridge to v3.1.0

v2.2.1 (13.05.2021)

🩹 Fixes:

  • 🐛 Fix: revert static plugin. It stays as a separate plugin on the main route (/) and supports all the previously announced features.
  • 🐛 Fix: remove build and other old targets from the Makefile.

v2.2.0 (11.05.2021)

👀 New:

  • ✏️ Reworked static plugin. Now, it does not affect the performance of the main route and persist on the separate file server (within the http plugin). Looong awaited feature: Etag (+ weak Etags) as well with the If-Mach , If-None-Match, If-Range, Last-Modified and If-Modified-Since tags supported. Static plugin has a bunch of new options such as: allow, calculate_etag , weak and pattern.

    Option always was deleted from the plugin.

  • ✏️ Update informer.List implementation. Now it returns a list with the all available plugins in the runtime.

🩹 Fixes:

  • 🐛 Fix: issue with wrong ordered middlewares (reverse). Now the order is correct.
  • 🐛 Fix: issue when RR fails if a user sets debug mode with the exec_ttl supervisor option.
  • 🐛 Fix: uniform log levels. Use everywhere the same levels (warn, error, debug, info, panic).

v2.1.1 (29.04.2021)

🩹 Fixes:

  • 🐛 Fix: issue with endure provided wrong logger interface implementation.

v2.1.0 (27.04.2021)

👀 New:

  • ✏️ New service plugin. Docs: link
  • ✏️ Stabilize kv plugin with boltdb, in-memory, memcached and redis drivers.

🩹 Fixes:

  • 🐛 Fix: Logger didn't provide an anonymous log instance to a plugins w/o Named interface implemented.
  • 🐛 Fix: http handler was without log listener after rr reset.

v2.0.4 (06.04.2021)

👀 New:

  • ✏️ Add support for linux/arm64 platform for docker image (thanks @tarampampam).
  • ✏️ Add dotenv file support (.env in working directory by default; file location can be changed using CLI flag --dotenv or DOTENV_PATH environment variable) (thanks @tarampampam).
  • 📜 Add a new raw mode for the logger plugin to keep the stderr log message of the worker unmodified (logger severity level should be at least INFO).
  • 🆕 Add Readiness probe check. The status plugin provides /ready endpoint which return the 204 HTTP code if there are no workers in the Ready state and 200 OK status if there are at least 1 worker in the Ready state.

🩹 Fixes:

  • 🐛 Fix: bug with the temporal worker which does not follow general graceful shutdown period.

v2.0.3 (29.03.2021)

🩹 Fixes:

  • 🐛 Fix: slow last response when reached max_jobs limit.

v2.0.2 (06.04.2021)

  • 🐛 Fix: Bug with required Root CA certificate for the SSL, now it's optional.
  • 🐛 Fix: Bug with incorrectly consuming metrics collector from the RPC calls (thanks @dstrop).
  • 🆕 New: HTTP/FCGI/HTTPS internal logs instead of going to the raw stdout will be displayed in the RR logger at the Info log level.
  • ⚡ New: Builds for the Mac with the M1 processor (arm64).
  • 👷 Rework ServeHTTP handler logic. Use http.Error instead of writing code directly to the response writer. Other small improvements.

v2.0.1 (09.03.2021)

  • 🐛 Fix: incorrect PHP command validation
  • 🐛 Fix: ldflags properly inject RR version
  • ⬆️ Update: README, links to the go.pkg from v1 to v2
  • 📦 Bump golang version in the Dockerfile and in the go.mod to 1.16
  • 📦 Bump Endure container to v1.0.0.

v2.0.0 (02.03.2021)

  • ✔️ Add a shared server to create PHP worker pools instead of isolated worker pool in each individual plugin.
  • 🆕 New plugin system with auto-recovery, easier plugin API.
  • 📜 New logger plugin to configure logging for each plugin individually.
  • 🔝 Up to 50% performance increase in HTTP workloads.
  • ✔️ Add Temporal Workflow plugin to run distributed computations on scale.
  • ✔️ Add debug flag to reload PHP worker ahead of a request (emulates PHP-FPM behavior).
  • ❌ Eliminate limit service, now each worker pool includes supervisor configuration.
  • 🆕 New resetter, informer plugins to perform hot reloads and observe loggers in a system.
  • 💫 Expose more HTTP plugin configuration options.
  • 🆕 Headers, static and gzip services now located in HTTP config.
  • 🆕 Ability to configure the middleware sequence.
  • 💣 Faster Goridge protocol (eliminated 50% of syscalls).
  • 💾 Add support for binary payloads for RPC (msgpack).
  • 🆕 Server no longer stops when a PHP worker dies (attempts to restart).
  • 💾 New RR binary server downloader.
  • 💣 Echoing no longer breaks execution (yay!).
  • 🆕 Migration to ZapLogger instead of Logrus.
  • 💥 RR can no longer stuck when studding down with broken tasks in a pipeline.
  • 🧪 More tests, more static analysis.
  • 💥 Create a new foundation for new KV, WebSocket, GRPC and Queue plugins.

v2.0.0-RC.4 (20.02.2021)

  • PHP tests use latest signatures (roadrunner-server#550).
  • Endure container update to v1.0.0-RC.2 version.
  • Remove unneeded mutex from the http.Workers method.
  • Rename checker plugin package to status, remove /v1 endpoint prefix (#557).
  • Add static, headers, status, gzip plugins to the main.go.
  • Fix workers pool behavior -> idle_ttl, ttl, max_memory are soft errors and exec_ttl is hard error.

v2.0.0-RC.3 (17.02.2021)

v2.0.0-RC.2 (11.02.2021)

  • Update RR to version v2.0.0-RC.2
  • Update Temporal plugin to version v2.0.0-RC.1
  • Update Goridge to version v3.0.1
  • Update Endure to version v1.0.0-RC.1