Implements the TaskQueue API.

For orchestrating the creation of the correct number of instances, managing errors and retries, and finally managing the deallocation of resources we use RabbitMQ in conjunction with the Celery Project framework, along with a self-developed workflow engine.

Task queues and cron jobs

백엔드, 어플리케이션 서버 모두 시간이 걸리는 작업들 및 주기적으로 수행해야 하는 작업들은 celery/celerybeat 을 통해 동작됩니다. 신뢰도가 아주 높으며, 유연합니다. gevent 로 동작시키면 더욱 깔끔한 관리가 가능합니다.

All of our background jobs (e.g., image resizing, file uploading, email and SMS sending) are done through Celery (using Redis as its broker). Celery's scheduling and retrying features are especially useful for error-prone tasks, such as email and SMS sending.

Using Celery, the web service creates tasks that are executed by a background worker. Celery uses a RabbitMQ instance as a task queue.

Automated distributed tasks.

We used celery, in combination with RabbitMQ and celery-beat, to run periodic tasks, as well as some user-initiated long-running tasks on the server.

Distributed Task Queue

万能胶也要省着用

http request 이후 바로 response가 즉시 오직 않는 많은 양의 무거운 작업은 celery에 task로 등록합니다.

그래서 유저에겐 빠른 response를 주고 celery에서 해당 업무를 수행합니다.

For offloading CPU intensive tasks to separate workers instead of running them on our application server.

background tasks.

python remote workers. simple and easy to use.

Task queues

Background Django/Python tasks. Populated Redis with server cache

Message passing.

Process queues

Primary message queue. Enqueueing operations revert to a local file-system-based queue when SQS is unavailable.

In the beginning we thought we wanted to start using something like RabbitMQ or maybe Kafka or maybe ActiveMQ. Back then we only had a few developers and no ops people. That has changed now, but we didn't really look forward to setting up a queuing cluster and making sure that all works.

What we did instead was we looked at what services Amazon offers to see if we can use those to build our own messaging system within those services. That's basically what we did. We wrote some clients in Ruby that can basically do the entire orchestration for us, and we run all our messaging on both SNS and SQS. Basically what you can do in Amazon services is you can use Amazon Simple Notification Service, so SNS, for creating topics and you can use queues to subscribe to these topics. That's basically all you need for a messaging system. You don't have to worry about scalability at all. That's what really appealed to us.

Message queue for our Elastic Beanstalk worker environments.

SQS is the bridge between our new Lambda services and our incumbent Rails applications. Extremely easy to use when you're already using other AWS infrastructure.

https://ifunny.co

Great pick-n-go queue, be prepared for occasionally outage in especially us-east-1 region.

Distributed background job queue.

I can't afford to lose data if Dynamo throttles my writes, so everything goes into a message queue first.

This isn't exactly low-latency (10s to 100s of milliseconds), but it has good throughput and a simple API. There is good reliability, and there is no configuration necessary to get up and running. A hosted queue is important when trying to move fast.

비디오 프로세싱 요청 메시지 큐
컨텐츠 배포 요청 메시지 큐

Message queue that the Raspberry Pi will retrieve messages.

For asynchronous data processing

Batch processing.

SQS is the primary means of communication between microservices.