https://yutongjin.com/tags/database/Recent content in Database on Yutong JinHugoen-usSun, 19 Apr 2026 00:00:00 +0000https://yutongjin.com/posts/419_learning/Sun, 19 Apr 2026 00:00:00 +0000https://yutongjin.com/posts/419_learning/<h2 id="relational-db-vs-nosql--interview-answer-template-with-qps-details">Relational DB vs NoSQL — Interview Answer Template (with QPS Details)</h2> <h3 id="1-throughput--scalability">1. Throughput &amp; Scalability</h3> <p>A relational database like MySQL can handle moderate QPS well.</p> <h4 id="rough-numbers-rule-of-thumb-varies-by-setup">Rough Numbers (rule of thumb, varies by setup):</h4> <ul> <li><strong>Single MySQL instance</strong> <ul> <li>~1K – 5K QPS (typical production range)</li> </ul> </li> <li><strong>With read replicas</strong> <ul> <li>Reads can scale to <strong>10K+ QPS</strong></li> </ul> </li> <li><strong>With sharding</strong> <ul> <li>Can scale to <strong>100K+ QPS</strong>, depending on number of shards</li> </ul> </li> </ul> <p>However, sharding introduces additional complexity:</p> <ul> <li>Choosing a good shard key</li> <li>Avoiding hotspot partitions</li> <li>Handling cross-shard queries</li> <li>Rebalancing data</li> </ul> <h4 id="example-tradeoffs">Example Tradeoffs:</h4> <ul> <li> <p><strong>Shard by userId</strong></p>