Top Tweets for #DATABases
3/ Peak insert throughput hit 1.55M rows/sec, and 20M rows stayed rock-solid at 64 MB RAM.
Next milestone: Write-Ahead Log (WAL) for crash recovery & ACID durability.
Code & benchmarks:
https://t.co/6a0BVyABMk
#buildinpublic #C #Databases
A 118M QPS benchmark tells you one thing: how fast single-row reads by key can go.
What's the most misleading benchmark you've seen?
#Databases #Postgres #SystemDesign
#AI #DBMS AI has lots of trouble debugging #databases. Even for a human DBA (database administrator), deadlocks and livelocks are trouble. Add a few triggers and stored procedures going sideways, and you must be smarter than Sherlock Holmes. I remember solving several of these issues where I had to think inside a topos (category theory) using three-valued #logic.... #risk
Where LLM Agents Get Stuck Debugging Databases: Access More Than Reasoning
Gustavo Pinto
https://t.co/4ISIjVKOtt [𝚌𝚜.𝚂𝙴]
💬Submitted to the Journal of Systems and Software
![ComputerPapers's tweet photo. Where LLM Agents Get Stuck Debugging Databases: Access More Than Reasoning
Gustavo Pinto
https://t.co/4ISIjVKOtt [𝚌𝚜.𝚂𝙴]
💬Submitted to the Journal of Systems and Software https://t.co/McGilfUqS2](https://pbs.twimg.com/media/HXEd4MHWwAA4XBA.png)
Merci, Montréal! 🇨🇦
Great conversations at @RedHat Summit: Connect about database licensing, infrastructure costs, and building more flexible, reliable systems.
Thanks to everyone who connected with us! 📸
#RHSummit #OpenSource #Databases #TechCommunity
Peter Zaitsev “DuckDB for #MySQL Old Timers”: what #DuckDB actually is, how it differs from MySQL, and how to start working with it.
DuckDB × MySQL Day, Nov 17, free online.
Register: https://t.co/eNcN6jMXfv
#opensource #databases #dba @PeterZaitsev @Percona @PerconaBytes
Design a ticket booking system: two people tap the last seat at the same moment. How do you make sure only one of them gets it?
#systemdesign #softwareengineering #codinginterview #backend #databases
My memory store needed every word of a search in 1 note. Paste a question, get nothing.
Now it ranks: any word counts, rarer words count more.
LoCoMo retrieval, recall@10: 39% to 59.5%. Plain BM25: 56.1%. p50 0.52 ms.
#SystemsEngineering #Databases #Rust
ProxySQL just added #DuckDB as a plugin, an embedded analytical database running right inside the proxy.
@rene_cannao (@proxysql ) explains it at DuckDB × #MySQL Day, Nov 17, free online.
Register: https://t.co/eNcN6jMXfv
#opensource #dba #sql #databases
Don't hold a database transaction open while waiting for a third-party API.
A slow HTTP call can leave locks open and keep a connection busy. Keep the transaction short, and decide how you'll record and retry the external step.
#Backend #Databases
To access the database, please click the link👇🏻
https://t.co/5F9pLaCbBM
#Lippincott
#Williams_and_Wilkins
#Databases
#Researcher_Services
Your signup validation can pass twice at the same time.
Two requests can both check that an email is free before either inserts it. Add a database unique constraint and handle the conflict in your API.
#Backend #SQL #Databases
Everyone says their solution is fast. Yevhen Patlan @Percona is testing that, with a benchmark comparing DuckDB-based solutions on the same 500 MB dataset.
#DuckDB × #MySQL Day, Nov 17, free online.
Register: https://t.co/eNcN6jMppX
#opensource #dba #databases @PerconaBytes
Six database engines. Nearly 1M embeddings. Indexing times from 1.9 minutes to 6.7 hours. 👀
Evgeniy Patlan compares vector search indexing, recall, storage and throughput—and the tradeoffs involved.
Results: https://t.co/fNQecmy9sO
#VectorSearch #AI #OpenSource #Databases
The same protein can appear under many different names across #databases, journals, and #patents. When #AI reads each name as a separate entity, it learns fragmented biology. Here is why drug discovery needs biology #ontologies: https://t.co/TreeSAp43w
Before writing a single line of storage code, I mapped out how SQLite actually moves data from a raw query string to disk.
Here is the exact architecture I’m implementing in C. Starting with the REPL & in-memory layer today! 🚀
#SystemsProgramming #Databases #C #BuildInPublic
Daniel Guzmán Burgos @clickslash (@dbtrail_ ) on replicating #MySQL into #Parquet and querying it with #DuckDB, no ETL, no warehouse.
DuckDB × MySQL Day, Nov 17, free online.
Register: https://t.co/eNcN6jMppX
#opensource #dba #databases #sql
Preallocated segments keep fsync to one disk round trip, but cost disk. Fresh AtticDB home was 130 MB (2 x 64 MiB segments). Now 16 MiB: 35.7 MB measured.
Floor is 16 MiB: a 4 MiB batch must fit in one. Create the spare on demand ?
https://t.co/gmJv1jJq6h
#Databases #Rust
TablePlus isn't the only option. Bdash is open source and self-hosted.
Bdash is a open-source databases. Modern SQL client application, supports MySQL, PostgreSQL (Redshift) and BigQuery.
→ Full sour...
#opensource #selfhosted #databases
Start with the query your users run. Then design the MongoDB index.
For “show this team’s newest open tickets,” try { teamId: 1, status: 1, createdAt: -1 }. The first two fields match the filters. The last supports newest-first sorting.
#MongoDB #Databases
Trends for you
Most Popular Users
Elon Musk 
@elonmusk
240.9M followers
Barack Obama
@barackobama
118.8M followers
Cristiano Ronaldo
@cristiano
114.5M followers
Donald J. Trump
@realdonaldtrump
111.9M followers
Narendra Modi
@narendramodi
107M followers
Rihanna
@rihanna
98.7M followers
NASA
@nasa
92.2M followers
Justin Bieber
@justinbieber
91.8M followers
KATY PERRY
@katyperry
90M followers
Taylor Swift
@taylorswift13
84M followers
Lady Gaga
@ladygaga
75.5M followers
Virat Kohli
@imvkohli
73.4M followers
Kim Kardashian
@kimkardashian
70.8M followers
YouTube
@youtube
68.7M followers
Neymar Jr
@neymarjr
66.4M followers
Bill Gates
@billgates
65.2M followers
Selena Gomez
@selenagomez
63.1M followers
The Ellen Show
@theellenshow
62.2M followers
CNN
@cnn
61.7M followers
X
@x
60.6M followers