Top all-time · #312
ScyllaDB
Company blog · scylladb.comdatasystems
ScyllaDB is the company blog at scylladb.com, writing about data and systems. Since 2015, 167 different Hacker News readers have linked it 424 times, most often “ScyllaDB: Drop-in replacement for Cassandra that claims to be 10x faster”. It is still writing: the last post was on 2026-10-29, and you can follow it by feed.
Rarely recommended · 3 of 40 sampled comments recommend it?
Why HN links it
HN readers mostly cite ScyllaDB's engineering blog for I/O scheduling, io_uring, benchmarks and coordinated omission, often as a source or to promote the product.
Summarised by Claude from a sample of 40 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Get unbreakable low-latency, high-throughput performance for massive AI workloads — from million-OPS feature stores to billion-scale vector search.
Most-linked posts?
- ScyllaDB: Drop-in replacement for Cassandra that claims to be 10x faster12 people · 12 mentions · first linked 2015-09-22
- Scylla – Real-Time Big Data Database10 people · 10 mentions · first linked 2018-03-05
- How io_uring and eBPF Will Revolutionize Programming in Linux8 people · 9 mentions · first linked 2020-05-10
- The Dark Side of MongoDB’s New License6 people · 9 mentions · first linked 2018-10-22
- Adventures with Memory Barriers and Seastar on Linux6 people · 6 mentions · first linked 2018-02-15
- Project Alternator: The Scylla Open Source Amazon DynamoDB-Compatible API6 people · 6 mentions · first linked 2019-09-11
- Discord Chooses Scylla as Its Core Storage Layer5 people · 6 mentions · first linked 2020-07-22
- Coordinated Omission (2021)5 people · 5 mentions · first linked 2021-09-29
- Scylladb: cassandra compatible NoSQL database5 people · 5 mentions · first linked 2015-09-23
- CockroachDB vs. Scylla Benchmark4 people · 5 mentions · first linked 2021-01-21
What Hacker News says?
> The JVM world tends to solve this problem by using off-heap caches. See Apache Ignite [0] or Ehcache [1]. Yeah, but I really do not bite your argument. When you are reduced to do manual memory management and fight the GC of your language, maybe you should simply not use a language with GC in the first place. They are right to use rust ( or C/C++) for that. It's not for nothing that redis (C) is so successful in th…
Comment linking /product/benchmarks/
on “Why Discord is switching from Go to Rust”Submitted How io_uring and eBPF Will Revolutionize Programming in Linux as a story · 709 points
HN discussionExactly. Regional distribution can be vital. Our customer Kiwi.com had a datacenter fire. 10 of their 30 nodes were turned to a slag heap of ash and metal. But 20 of 30 nodes in their cluster were in completely different datacenters so they lost zero data and kept running non-stop. This is a rare story, but you do NOT want to be one of the thousands of others that only had one datacenter, and their backups were also…
Comment linking /2021/03/23/kiwi-com-nonstop-operations-with-scylla-even-through-the-ovhcloud-fire/
on “Use one big server”Submitted I/O Access Methods for Linux as a story · 268 points
HN discussionWe at Scylla tune the kernel, drives, stride, clock source automatically and benchmark the disk to figure out what's its maximum parallelism so it won't queue (we can saturate 8 NVMe) and we'll do the queuing within the DB's userspace so we can prioritize real time queries over background tasks: https://www.scylladb.com/2018/04/19/scylla-i-o-scheduler-3/
Comment linking /2018/04/19/scylla-i-o-scheduler-3/
on “I/O Is Faster Than CPU – Let’s Partition Resources and Eliminate OS Abstractions [pdf]”Submitted Is Arm ready for server dominance? as a story · 251 points
HN discussionI lol'ed. (Full disclosure: I work for ScyllaDB.) MongoDB has definitely come a long way in terms of HA, but yes, they are still have a long way to go. A good primer talk on the differences can be found here: https://www.scylladb.com/tech-talk/mongodb-vs-scylla-product...
Comment linking /tech-talk/mongodb-vs-scylla-production-experience-from-both-dev-ops-standpoint/
on “Jepsen: PostgreSQL 12.3”Submitted Async Rust in Practice: Performance, Pitfalls, Profiling as a story · 248 points
HN discussion> Scylla did a C++ rewrite to address tail latency due to JVM GC, but after an explosive release cycle, they basically stalled at partial 2.2 compatiblity Could you explain this more? Because Scylla has had pretty steady major release updates over the past few year. See the timeline of updates here: https://www.scylladb.com/2021/08/24/apache-cassandra-4-0-vs-... We have long since passed C* 3.11 compatibility. In fa…
Comment linking /2021/08/24/apache-cassandra-4-0-vs-scylla-4-4-comparing-performance/
on “How Discord Stores Billions of Messages (2017)”Submitted ScyllaDB Open Source 3.0 as a story · 210 points
HN discussionThat sounds like one of my favorite use cases, Meshify / MunichRE! :) A little history lesson is in order: https://www.scylladb.com/2019/02/01/meshify-and-scylla-an-in...
Comment linking /2019/02/01/meshify-and-scylla-an-industrial-strength-iot-solution/
on “How Discord Stores Billions of Messages (2017)”Submitted New ScyllaDB Go Driver: Faster Than GoCQL and Its Rust Counterpart as a story · 183 points
HN discussionIf you are doing north of a million ops on DynamoDB you can quickly run into the $2-3 million a year range. In this 2018 benchmark, we were able to calculate that a sustained, provisioned of only 160k write ops / 80k read ops for DynamoDB would cost >$500k per year: https://www.scylladb.com/2018/12/13/scylla-vs-amazon-dynamod... That was a few years ago. These days, according to our most current pricing you could do…
Comment linking /pricing/?writes=500000&reads=500000&storage=1&itemSize=1&replication=3
on “How Discord Stores Billions of Messages (2017)”Submitted How Palo Alto Networks Replaced Kafka with ScyllaDB for Stream Processing as a story · 126 points
HN discussionWe just benchmarked Cassandra 4.0, which is brand-spanking new. The good news: C 4.0 is a far better performing database than C 3.11. The new GCs definitely get rid of the long tail latency nightmares: https://www.scylladb.com/2021/08/19/cassandra-4-0-vs-cassand... However, we also compared it to Scylla's latest release, and though C 4 is better*, you can still find other CQL-compatible databases that outperform it.…
Comment linking /2021/08/24/apache-cassandra-4-0-vs-scylla-4-4-comparing-performance/
on “How Discord Stores Billions of Messages (2017)”Submitted Making Faster B-Trees with Go Generics as a story · 119 points
HN discussionThis would not work at scale for a company like Discord, with its volume of traffic. Cockroach, being consistency-oriented would quickly become transaction-bound. You want a database like a Cassandra or Scylla that is more performance/availability oriented. Otherwise you are going to see a lot of lag and latency in the Discord chat. Cockroach is very, very good for a distributed SQL database. But it's still performa…
Comment linking /2021/01/21/cockroachdb-vs-scylla-benchmark/
on “How Discord Stores Billions of Messages (2017)”Submitted ScyllaDB: Drop-in replacement for Cassandra that claims to be 10x faster as a story · 114 points
HN discussionThey already did— https://news.ycombinator.com/item?id=28293097 , main article: https://www.scylladb.com/press-release/discord-chooses-scyll...
Comment linking /press-release/discord-chooses-scylla-core-storage-layer/
on “How Discord Stores Billions of Messages (2017)”Submitted Scylla – Real-Time Big Data Database as a story · 96 points
HN discussion...then [post this blog] Scylla. https://www.scylladb.com/press-release/discord-chooses-scyll...
Comment linking /press-release/discord-chooses-scylla-core-storage-layer/
on “How Discord Stores Billions of Messages (2017)”Submitted Designing a Userspace Disk I/O Scheduler for Modern Datastores (2016) as a story · 91 points
HN discussionYes. They started experimenting with Scylla earlier, and made the switch in 2020. Here's more on their logic — they like "opinionated systems": https://www.scylladb.com/2019/03/20/discord-on-the-joy-of-op...
Comment linking /2019/03/20/discord-on-the-joy-of-opinionated-systems/
on “How Discord Stores Billions of Messages (2017)”Submitted ScyllaDB – Why We're Moving to a Source Available License as a story · 82 points
HN discussionLooks like they migrated (at least partially) to Scylla: "Discord Chooses Scylla as Its Core Storage Layer" (2020) https://www.scylladb.com/press-release/discord-chooses-scyll...
Comment linking /press-release/discord-chooses-scylla-core-storage-layer/
on “How Discord Stores Billions of Messages (2017)”Submitted Going Head-To-Head: Scylla vs. Amazon DynamoDB as a story · 77 points
HN discussionCorrect. Redpanda is source-available. When you have C++ code, the number of external folks who want to — and who can effectively, actively contribute to the code — drops considerably. Our "cousins in code," ScyllaDB last year announced they were moving to source-available because of the lack of OSS contributors: > Moreover, we have been the single significant contributor of the source code. Our ecosystem tools have…
Comment linking /2024/12/18/why-were-moving-to-a-source-available-license/
on “Kafka is Fast – I'll use Postgres”Submitted ScyllaDB Closes $16M in Series B Funding as a story · 76 points
HN discussionIn the article, it states that after a certain point, parallelization hits limits. Is that due to Scylladb's architecture or other constraints within the system? For reads (unless there is another bottleneck), by 10x'ing the parallelism, an application can compensate for 10x the average latency and still deliver the same throughput. Scylladb talked about their IO scheduler a good deal: https://www.scylladb.com/2021/…
Comment linking /2022/08/03/implementing-a-new-io-scheduler-algorithm-for-mixed-read-write-workloads/
on “How Discord supercharges network disks for extreme low latency”Submitted Designing the Scylla Userspace Disk I/O Scheduler (Part 1) as a story · 72 points
HN discussion
