---
title: "Chief Debugger (@sudhirr) | Hashnode"
description: "Chief Debugger (@sudhirr) writes on Hashnode. Read their latest articles and follow their work."
url: "https://hashnode.com/@sudhirr"
---

@sudhirr

# Chief Debugger

@sudhirr · Bangalore, India · Joined April 2026

### [The Bug That Only Showed Up After a Week: Why We Still Needed to Add a Daily VACUUM](https://tech.kompress.ai/the-bug-that-only-showed-up-after-a-week-why-we-still-needed-to-add-a-daily-vacuum?utm_source=hashnode&utm_medium=feed)

Part 3 of 3 — A Quieter Story About Ongoing Ownership

6 min read · [tech.kompress.ai](https://tech.kompress.ai/the-bug-that-only-showed-up-after-a-week-why-we-still-needed-to-add-a-daily-vacuum?utm_source=hashnode&utm_medium=feed)

[Jul 8](https://hashnode.com/posts/the-bug-that-only-showed-up-after-a-week-why-we-still-needed-to-add-a-daily-vacuum/6a4db3353766825fd0caaaed)

### [Part 3: Building a Cluster That Manages Itself](https://tech.kompress.ai/part-3-building-a-cluster-that-manages-itself?utm_source=hashnode&utm_medium=feed)

The partition index table, the monitoring system, and reading the sawtooth

10 min read · [tech.kompress.ai](https://tech.kompress.ai/part-3-building-a-cluster-that-manages-itself?utm_source=hashnode&utm_medium=feed)

[Jun 20](https://hashnode.com/posts/part-3-building-a-cluster-that-manages-itself/6a36c565cbd528a13178c762)

### [Part 2: When 0% Tombstones Is Not the Finish Line"](https://tech.kompress.ai/part-2-when-0-tombstones-is-not-the-finish-line?utm_source=hashnode&utm_medium=feed)

Four days after hitting 0.00000% droppable tombstones on all three nodes simultaneously — the cleanest state the cluster had ever been in — the Slack report showed 30% again

8 min read · [tech.kompress.ai](https://tech.kompress.ai/part-2-when-0-tombstones-is-not-the-finish-line?utm_source=hashnode&utm_medium=feed)

[Jun 19](https://hashnode.com/posts/part-2-when-0-tombstones-is-not-the-finish-line/6a34ef3ccea6d0ce89b1c364)

### [Part 1: When Your Time-Series Data Turns Against You...](https://tech.kompress.ai/part-1-when-your-time-series-data-turns-against-you?utm_source=hashnode&utm_medium=feed)

Migrating 3.9 Billion Rows to TWCS Without Downtime on a Constrained IoT Cluster The number that started everything was 7.99 ms. That was the average read latency on our Cassandra keyspace — a cluster

11 min read · [tech.kompress.ai](https://tech.kompress.ai/part-1-when-your-time-series-data-turns-against-you?utm_source=hashnode&utm_medium=feed)

[Jun 15](https://hashnode.com/posts/part-1-when-your-time-series-data-turns-against-you/6a2f48a3c68290808f39efc8)

### [When Your Fix Creates a New Problem: Debugging the Production Partition Migration](https://tech.kompress.ai/when-your-fix-creates-a-new-problem-debugging-the-production-partition-migration?utm_source=hashnode&utm_medium=feed)

Part 2 of 2 — Two Swap Cycles, One Rollback, and Why the Safety Net Made All the Difference

9 min read · [tech.kompress.ai](https://tech.kompress.ai/when-your-fix-creates-a-new-problem-debugging-the-production-partition-migration?utm_source=hashnode&utm_medium=feed)

[1](https://hashnode.com/posts/when-your-fix-creates-a-new-problem-debugging-the-production-partition-migration/6a2a0aadac70f47cc6b87078) [Jun 11](https://hashnode.com/posts/when-your-fix-creates-a-new-problem-debugging-the-production-partition-migration/6a2a0aadac70f47cc6b87078)
