Jatin Jain Saraf·Oct 27Bloom Filter, Reservoir Sampling, HyperLogLog: Three Probabilistic Answers to Three Different…Same family, different jobsA response icon1A response icon1
Jatin Jain Saraf·Oct 22Your Node.js App Never Imports zlib. It Might Still Be Vulnerable to a Brotli RCEOn October 21, 2026, Node.js shipped security releases for all three supported lines. The headline fix is CVE-2026–111497, a High-severity…
Jatin Jain Saraf·Oct 22You Might Not Need a Queue Service: Postgres SKIP LOCKED as a Job QueueThe Problem Every Indexer Hits Eventually
Jatin Jain Saraf·Aug 29The Dual-Write Problem: Why “Update the DB, Then Publish an Event” Is BrokenAn order gets saved to Postgres. The next line of code publishes an “order created” event to a message broker, so the payments service can…
Jatin Jain Saraf·Aug 15Why Postgres and Cassandra Made Opposite Bets on Storage EnginesA write-heavy ingestion table starts slow on Postgres. Not immediately, but a few weeks in: inserts that used to take a millisecond now…A response icon1A response icon1
Jatin Jain Saraf·Jul 27PostgreSQL 19 vs. 18: What Actually Changed, Feature by FeatureEvery major PostgreSQL release gets the same LinkedIn headline treatment: “this changes everything.” Most of the time it doesn’t, it’s a…
Jatin Jain Saraf·Jul 21How the LRU Cache Actually Works and Why Redis, Browsers, and Postgres All Approximate ItYou’ve felt this before: you restart a service, or open your laptop after the weekend, and everything is sluggish for the first few…
Jatin Jain Saraf·Jul 15Redis HyperLogLog: Counting Millions of Unique Users with Just 12KB of MemoryWhen building at scale, answering a seemingly simple question like “How many unique users visited our site today?” can become a massive…
Jatin Jain Saraf·Jul 15Why a Single ALTER TABLE Can Take Down Your Whole DatabaseA team runs a routine migration: ALTER TABLE transactions ADD COLUMN region TEXT. Nothing dramatic on paper. Twelve minutes later the…
Jatin Jain Saraf·Jul 14Your Blockchain Indexer Is Fine — Until It Isn’tWhat nobody tells you about hitting 20–30 TB of indexed blockchain data, and why “just optimize it later” is the most expensive sentence…