<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Confluent on </title>
    <link>/tags/confluent/</link>
    <description>Recent content in Confluent on </description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en</language>
    <lastBuildDate>Mon, 30 Mar 2026 10:04:30 +0800</lastBuildDate><atom:link href="/tags/confluent/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Confluent Kafka 业务分区数量评估笔记</title>
      <link>/posts/confluent-kafka-partition-count-evaluation/</link>
      <pubDate>Mon, 30 Mar 2026 10:04:30 +0800</pubDate>
      
      <guid>/posts/confluent-kafka-partition-count-evaluation/</guid>
      <description>背景 我们的项目使用 Confluent Cloud 托管 Kafka 集群，在上线前需要确定每个 topic 的分区（partition）数量。Kafka 的分区数只能增加不能减少，因此更适合在初始阶段先做一轮带假设前提的评估。
本文记录这一评估过程。
业务规模 基础数据 (单 Topic) 参数 数值 单 Topic 日均消息量 4 亿条 活跃用户数 ~200 万 人均日均交易笔数 200 笔 每笔交易产生消息数 1 条 Topic 数量 约 60 个 集群日均总消息量 约 240 亿条 单 Topic RPS 推导 单 Topic 全天平均 RPS = 4亿 ÷ 86,400秒 ≈ 4,630 交易流量并非均匀分布。假设流量集中在约 8 小时的活跃窗口内：
单 Topic 活跃时段平均 RPS = 4亿 ÷ (8 × 3,600秒) ≈ 13,900 活跃窗口内还存在早晚高峰，峰值通常为活跃时段均值的 2–3 倍：</description>
    </item>
    
    <item>
      <title>How Many Kafka Connections Does Your Spring Boot App Actually Use?</title>
      <link>/posts/confluent-kafka-connection-count-spring-boot/</link>
      <pubDate>Thu, 12 Mar 2026 06:00:00 +0800</pubDate>
      
      <guid>/posts/confluent-kafka-connection-count-spring-boot/</guid>
      <description>When selecting a Confluent Cloud plan, one of the first limits you&amp;rsquo;ll hit is the maximum connection count. Enterprise clusters allow 18,000 connections per eCKU (raised 4× in the Q3 2025 update); Standard cluster limits are lower — check the current cluster types table for the exact figure. Miscounting by even one component can push you into throttling territory.
This post gives you a practical formula for estimating connections in a Spring Boot application, explains what actually drives that number, and ends with a worked example for a multi-service deployment.</description>
    </item>
    
  </channel>
</rss>
