---
title: "Lessons learned building DoorDash’s clusterless ML feature store"
url: https://daily.dev/posts/lessons-learned-building-doordash-s-clusterless-ml-feature-store-wbskyh8za
source_url: https://careersatdoordash.com/blog/doordash-clusterless-ml-feature-store
type: article
source: "Doordash"
published: 2026-05-18T16:29:49.190Z
updated: 2026-05-18T16:31:09.892Z
tags: ["distributed-systems", "redis"]
reading_time: 10
upvotes: 1
comments: 0
language: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Lessons learned building DoorDash’s clusterless ML feature store

**[Doordash](https://daily.dev/sources/doordash)** · 10 min read · 1 upvotes · 0 comments

## Summary

DoorDash shares how they built a clusterless, stateless ML feature store to replace a costly Redis-plus-relational-database hybrid that couldn't scale further. The new system uses Apache Kvrocks (a Redis-protocol-compatible store backed by RocksDB) running on commodity SSD instance-store disks, with a custom Redis Cluster Manager (RCM) that provides topology transparency to clients without requiring actual cluster state sharing. Data is batch-loaded from S3 as RocksDB backups, enabling stateless horizontal scaling. A two-phase rollout (shadow validation then traffic migration) validated the design. A key discovery was that large Redis clusters degrade client performance at 2,000+ nodes, solved by having RCM return a per-client subset of nodes. The system now handles 130M HMGETs per second serving 1.6B features within a 50ms P999 latency target.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://careersatdoordash.com/blog/doordash-clusterless-ml-feature-store>

## Similar posts on daily.dev

- [Rediscovering RocksDB – Embedded Storage in Cloud-Native Applications](https://daily.dev/posts/rediscovering-rocksdb-embedded-storage-in-cloud-native-applications-mfiffjtev) · Cloud Native Now · 1 upvotes · 0 comments
- [CockroachDB Serverless: Sub-second Scaling from Zero with Multi-region Cluster Virtualization](https://daily.dev/posts/cockroachdb-serverless-sub-second-scaling-from-zero-with-multi-region-cluster-virtualization-3eawwrikn) · Metadata · 0 upvotes · 0 comments
- [How Databricks Feature Store serves features with sub-second freshness](https://daily.dev/posts/how-databricks-feature-store-serves-features-with-sub-second-freshness-shhczz1nn) · databricks · 0 upvotes · 0 comments
- [Moving Marketing Cloud Caching from Memcached to Redis at 1.5M RPS](https://daily.dev/posts/moving-marketing-cloud-caching-from-memcached-to-redis-at-1-5m-rps-epsn0tifr) · Salesforce Engineering · 1 upvotes · 0 comments

---

Tags: [#distributed-systems](https://daily.dev/tags/distributed-systems), [#redis](https://daily.dev/tags/redis)

[View this post on daily.dev](https://daily.dev/posts/lessons-learned-building-doordash-s-clusterless-ml-feature-store-wbskyh8za)