@problame

Software engineer in the storage systems domain. Working @databricks on @neondatabase. Mastodon: @problame@dgc.social Bluesky: @problame.bsky.social

 Karlsruhe, Germany
cschwarz.com
Joined November 2009
  • Tweets5.2K
  • Following612
  • Followers435
168Media
Arrived in Portland for @TokioConf . I’m speaking Wednesday about when and how io_uring can make your storage-intensive tokio app better. Check out the full schedule, awesome lineup! tokioconf.com/schedule
1
1
7
1,934
Super excited to be speaking at #p99conf about the evolution of the Neon storage stack! My talk is on Wednesday: p99conf.io/agenda/#wednesday

P99 Conference | ScyllaDB

The event for developers who care about P99 percentiles and high-performance, low-latency applications.

p99conf.io
P99CONF
@P99CONF29 Sep 2025
A nice range of distributed data systems, approaches, and challenges are represented at #P99CONF. Check out this sneak peek of @qianl_cs, @avikivity, @glcst, @NickVanWig, @problame, @nehapawar18, @SivukhinN, @nyharel, and @andy_pavlo's sessions. ow.ly/6a8E50X3P8l #ScyllaDB
1
3
348
And I just received this amazing swag drop in the mail: that beanie is going for a morning run right away!
1
113
Replying to @Microsoft
@Microsoft announces @neondatabase first party service on @Azure
1
3
32
1,042
Christian Schwarz retweeted
Neon
@neondatabase7 Nov 2024
Ever wondered how Neon's autoscaling actually works? Check out this technical deep dive which covers how we make Postgres on Neon adapt to your workload ↓
1
3
2
18
3,426
Christian Schwarz retweeted
Neon
@neondatabase5 Sep 2024
As part of Autoscaling GA, we made a major improvement to the algorithm: We now estimate working set size and autoscale to keep it in memory. Result: up to 8x improvement in throughput on read-heavy workloads. Read about how it works here: neon.tech/blog/dynamically-e…

Dynamically estimating and scaling Postgres’ working set size - Neon

A deep dive at the technical details of how Neon estimates the Postgres working set size to keep the working set in memory via autoscaling.

neon.com
7
21
3,988
Christian Schwarz retweeted
Neon
@neondatabase15 Apr 2024
⚡Storage Sharding ⚡️ Neon rethinks Postgres storage to make it more scalable 🐘 We implement transparent storage sharding in our engine to support larger storage capacity and throughput For devs, it's invisible but here's how we do it 👇 🧵
2
14
2
102
21,735
Christian Schwarz retweeted
Neon
@neondatabase15 Apr 2024
Neon is now GA 🎉🎉🎉
Neon Postgres is now GA

ALT Neon Postgres is now GA

11
64
29
336
94,173
John Spray, Storage Engineering Lead @neondatabase, shares the complexities of stateful service design in the #cloud with their Neon Serverless Postgres case study shorturl.at/gkoUY #QConLondon #CloudNative #CloudNativeEngineering #CloudEngineering #SoftwareDevelopment
3
11
1,337
System engineering work done @neondatabase. Technical. neon.tech/blog/posix_spawn-c…

How posix_spawn & close_range "fixed" cold starts - Neon

Last November, we had several outages and have since directed much of our engineering efforts towards improving stability. During this phase, cold start latency regressed, but the root cause and...

neon.com
1
3
10
4,924
jack-vanlightly.com/analyses… - @vanlightly did a really good overview of the @neondatabase architecture and rightfully contrasted it with AWS Aurora.

Neon - Serverless PostgreSQL - ASDS Chapter 3 — Jack Vanlightly

Neon is a serverless Postgres service based on an architecture similar to Amazon Aurora . It separates the Postgres monolith into disaggregated storage and compute. The motivation behind this...

jack-vanlightly.com
1
31
3,209
Christian Schwarz retweeted
VentureBeat
@VentureBeat1 Aug 2023
Neon raises $46 Million to advance serverless PostgreSQL database for the AI era venturebeat.com/data-infrast…

Neon raises $46M for serverless PostgreSQL | VentureBeat

Neon offers a serverless cloud service, providing the open-source PostgreSQL (also known as 'Postgres') relational database.

venturebeat.com
3
9
3,730
Very excited to launch the second puzzle piece after separating storage for @neondatabase. Autoscaling with live VM migrations. All code is opensourced! neon.tech/blog/postgres-auto…

Scaling serverless Postgres: How we implement autoscaling - Neon

We’re excited to announce the release of autoscaling in Feature-Preview! This feature allows your database to accommodate varying workloads and meet your application’s demand. In this blog post,...

neon.com
3
6
1
51
4,351
Also: macOS's native NTFS implementation has a (similar?) bug where dir entries with umlauts simply don't show up in Finder, if these umlauts are encoded in an unexpected way.
Christian Schwarz@problame5 Feb 2023
TIL: the built-in macOS SMB client does implicit unicode normalization of any file names in the requests it sends. To NFC-form, IIUC.
2
2
572
The effect is frightening: drag&drop of a directory that contains such entries will silently skip over them. rsync will notice it and print a `vanished directory` error.
1
1
259
Used an NTFS-formatted external HDD to migrate some files to a Mac in the past? You might have lost data if you simply did drag&drop. Imagine the scale of this. Also, I'm wondering if the same bug exists for FAT (=> USB flash drives).
1
4
261
TIL: the built-in macOS SMB client does implicit unicode normalization of any file names in the requests it sends. To NFC-form, IIUC.
1
1
1
877
more replies
"Fixing" this: 1. rewrite all dir entries to macOS's normalization form (NFC, IIUC), and 2. configure the SMB server and/or the server-side filesystem to force that normalization for all clients. Obviously (2) impacts clients that don't normalize implicitly, e.g., Linux CIFS.
1
1
185
The best write-up I found _after_ figuring this out: truenas.com/docs/scale/scale…
154
Load more