news.volyx.in

BlazingMQ: High-performance open source message queuing system (bloomberg.github.io)

695 points by carride · 1126 days ago · 259 comments on HN

Article summary

BlazingMQ is a distributed message queueing platform designed for efficiency, reliability, and a rich feature set. It provides consistently low median and p99 latency, and its unique multi-hop network topology can lead to significant savings in network bandwidth and latency. The platform is built on industry-standard best practices in distributed systems and offers a range of features, including message routing strategies and clustering. BlazingMQ is open-source and licensed under Apache 2.0.

Main themes

  • message queueing
  • distributed systems
  • performance optimization
  • software design
  • animation and visualization
  • technical communication
  • system architecture
  • scalability and reliability

What commenters say

  • The landing page for BlazingMQ is well-designed and effectively communicates the platform's features and benefits.
  • The use of animations on the landing page is a useful tool for understanding complex technical concepts.
  • Some commentators question the need for another message queueing system, given the existence of established alternatives like RabbitMQ and Kafka.
  • The performance of BlazingMQ is a key differentiator, with some commentators noting that its ability to handle high volumes of messages is a significant advantage.
  • Others argue that the concept of a 'high-performance' message queue is misleading, as queueing is inherently a bottleneck.
  • The choice of server count in a distributed system is a complex issue, with different approaches having different trade-offs in terms of performance, reliability, and cost.
  • Some commentators suggest that BlazingMQ's newness may be an advantage, as it can take advantage of recent advances in distributed systems and programming languages.
  • The use of message queues can actually improve system performance by allowing for buffering and asynchronous processing, rather than blocking on individual requests.