내가 좋아할 만한 콘텐츠
When you build a database, edge cases become part of your daily routine. From concurrency challenges to timestamp precision, we hit them all. At #QuestDB, that’s just what a productive Tuesday looks like.
Did you know #QuestDB service accounts never have passwords? You log in with your user or SSO, then use ASSUME SERVICE ACCOUNT to act as one, and EXIT SERVICE ACCOUNT to switch back. A simple and secure way to run automation or delegated tasks.
When testing data patterns or query performance, uniform randomness isn’t always realistic. The new rnd_symbol_zipf() in #QuestDB lets you generate skewed synthetic datasets, with or without weights, for your symbol columns. Perfect for real-world performance testing.
We just merged gzip support for ILP workloads in #QuestDB. Clients can now compress data before sending, reducing bandwidth usage and boosting throughput performance. Smaller payloads, faster ingestion. ⚡
We’re cooking something new for #QuestDB. 🍳 You’ll soon be able to use your own Claude or OpenAI API keys right from the Web Console. Ask in plain text, get SQL and results back. It can also help debug errors or explain queries. Screenshots 👇
How do you say in your database “for each row in A, find the closest record in time from table B matching by symbol, but only if it’s equal or earlier by no more than 100 nanoseconds”? In #QuestDB, it’s one elegant SQL query. 👇
With full outer joins in #QuestDB, you no longer need to stitch results together with two left joins and a UNION. One query does it all now, clean and simple.
If you’re riding the London Tube these days, look up. You might see what #QuestDB is about. Our new campaign celebrates performance and openness, with support for formats like Parquet that keep your data portable and vendor lock-free.
The new timestamp_ns data type in #QuestDB enables nanosecond precision analytics. Perfect for downsampling or ASOF JOINs with nanosecond tolerance, for those cases where missing the exact timestamp is not an option.
We were casually running ClickBench on a 32-core desktop and noticed #QuestDB got faster. Ingest speed is now 5.9M rows/sec, about 100K/sec more than our last run.
You already know how to make highly available reads in #QuestDB. Coming soon to our official ILP clients: highly available writes using multiple node names with automatic retries.
Floats and doubles are fast, but not always precise. Sometimes that matters a lot. Coming soon to #QuestDB: a new DECIMAL type for exact math. Perfect when every digit counts.
How many database developers do you need to fix a race condition bug? 🤔 #QuestDB says: SELECT rnd_int(1, 10, 0);
United States 트렌드
- 1. D’Angelo 176K posts
- 2. D’Angelo 176K posts
- 3. Brown Sugar 14.8K posts
- 4. Black Messiah 6,952 posts
- 5. Voodoo 14.6K posts
- 6. #GoodTimebro 1,458 posts
- 7. Happy Birthday Charlie 118K posts
- 8. How Does It Feel 6,789 posts
- 9. Powell 36K posts
- 10. #PortfolioDay 9,360 posts
- 11. Osimhen 110K posts
- 12. #BornOfStarlightHeeseung 80.6K posts
- 13. Pentagon 96.4K posts
- 14. Neo-Soul 15.9K posts
- 15. Sandy Hook 10.8K posts
- 16. Alex Jones 27.1K posts
- 17. Young Republicans 3,292 posts
- 18. Jill Scott 1,087 posts
- 19. Untitled 5,951 posts
- 20. CJGJ N/A
내가 좋아할 만한 콘텐츠
-
DuckDB
@duckdb -
Fly.io
@flydotio -
Apache Pulsar
@apache_pulsar -
Dragonfly
@dragonflydbio -
ClickHouse
@ClickHouseDB -
Debezium Project
@debezium -
Timescale (now TigerData)
@TimescaleDB -
TiDB, powered by PingCAP
@PingCAP -
Apache Pinot
@ApachePinot -
InfluxData
@InfluxDB -
Milvus
@milvusio -
Apache Doris
@doris_apache -
Grafbase
@grafbase -
Chainguard ⛓️
@chainguard_dev -
RS Rajakannappan- தமிழ்நாட்டை தலைகுனிய விடமாட்டேன்
@RRajakannappan
Something went wrong.
Something went wrong.