Vinoth Chandar @byte_array

Founder @Onehousehq — Data Infra for Agents. Creator of @apachehudi, built the first #DataLakehouse @Uber. Databases, Dist. Systems. LinkedIn, Confluent. bytearray.io Joined April 2009

Tweets

1K
Followers

2K
Following

235
Likes

831

Vinoth Chandar @byte_array

2 days ago

@ClaudeDevs This has got to be the most hype that gzip has had in a decade or more.

0 0 0 134 0

View Details

Vinoth Chandar @byte_array

4 days ago

Alright . That is in fact pretty freakin' cool.

Benny Johnson @bennyjohnson

4 days ago

🚨 The new Tesla Roadster can FLY. This is INSANE. Only Elon Musk and only in America 🚀

598 1K 12K 1.5M 1K

0 0 0 173 0

View Details

Vinoth Chandar @byte_array

4 days ago

@rabata_io I feel this is AI generated. 🙂 What tool are you using?

0 0 0 11 0

View Details

Last week, I posted on why warehouses still default to closed formats. What I learned: they won't flip the default. Open formats are opt-in. So the fix is with the users. Give every open format (Delta, Hudi, Lance, Vortex, ClickHouse, DuckLake) the same consideration on performance and flexibility that you give the closed ones. Your data will thank you. Back to building on open formats.

1 1 5 258 0

View Details

Vinoth Chandar @byte_array

6 days ago

@rmahindra Does the technical problem simply boil down to a very large table on object storage with fast writes and fairly fast analytical and point lookups?

0 0 0 55 0

View Details

Vinoth Chandar @byte_array

7 days ago

I like that the ChatGPT Mac app has a built-in markdown editor. It definitely reduces friction in review cycles.

3 0 5 311 0

View Details

Vinoth Chandar @byte_array

a week ago

@jaykreps My wrists are happy that I have to type less. This feeling might be same elsewhere too

1 0 0 196 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

Not at all. Customers are stuck. We speak to dozens every week. They have to pay hefty egress costs to do the migration. The question I raised is more about first principle and forward-looking. Why not make open formats the default for any new table created from this point forward? That costs the vendor nothing. Btw the LinkedIn counterpart for this x post has a lot more thoughts if you are curious about this topic .

0 0 1 30 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

Last week, Databricks CEO Ali Ghodsi and I debated Iceberg and open table formats on a post. Then I caught myself. Why are we debating each other while Snowflake, BigQuery and Redshift still default customers to closed formats? Two years of Iceberg announcements. Still no public retirement date for their closed formats. Open-source projects get scrutinized for interoperability on every release. The largest warehouses should meet the same bar. We’ve spent a decade building open formats and data lakes. I think it’s time we asked.

3 3 13 749 2

View Details

Vinoth Chandar @byte_array

2 weeks ago

My thoughts and receipts are in the blog onehouse.ai/blog/snowflake…

0 0 0 67 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

@rabata_io lol it's actually pretty basic when you see your servers only have 50% CPU utilization.

0 0 0 11 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

To squeeze the performance out of our runtime components, we implemented a good amount of concepts from - SEDA: an architecture for well-conditioned, scalable internet services dl.acm.org/doi/10.1145/50… The seminal work at MIT around writing high-performance server software and the gift that keeps on giving.

Onehouse @Onehousehq

2 weeks ago

A cloud VM’s network I/O, CPU, and storage IOPS taking turns instead of running concurrently is bad news: the CPU sits idle while a file downloads, then works through it once the download finishes.

1 0 2 422 0

1 1 3 376 1

View Details

Vinoth Chandar @byte_array

2 weeks ago

@shcallaway @sazabi I don't have extreme opinions on unit tests, but I hate the false sense of security they give developers, who only later realize the end-to-end functionality doesn't work at all.

0 0 1 76 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

Muse, Dots, Grok bot. Eagerly awaiting the consumer AI/agent app explosion, to watch agents melt our databases at scale.

0 0 2 205 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

Impressive numbers.

Daniel Lemire @lemire

3 weeks ago

You can consume compressed JSON documents in a multithreaded manner if you use zstd for compression. This allows crazy speeds. See my article.

6 30 331 20K 183

0 0 3 401 1

View Details

Vinoth Chandar @byte_array

2 weeks ago

@samuel_spitz What if you need a super innovative idea to collect the $0.01 ?

0 0 0 201 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

@a16z Consumer lives and livelihood improving will the ultimate test for this AI wave? Once we start to run out of ideas to use AI to make enterprises efficient.

0 0 0 63 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

@CorvusCrypto Not really. more like tradeoffs around madness

0 0 0 163 0

View Details

Vinoth Chandar @byte_array

3 weeks ago

🙄 This is just confusing at so many levels. First: historical data is several times larger than live data. How would a single-node SQL engine crunch all that data volume? Amazon itself sells three distributed SQL engines, last time I checked Second: you can easily have your agent hit both Aurora and Athena and solve this problem today. What's broken?

Andy Jassy @ajassy

3 weeks ago

A painful part of working with your data has always been that your live data and historical data are stuck in separate systems: the order a customer just placed lives in your database, while their last five years of orders sit in a data lake in S3. And answering a real question

144 151 2K 629K 1K

5 1 18 9K 7

View Details

Vinoth Chandar @byte_array

2 weeks ago

@mr_lazy_coder We built a columnar caching layer with in-memory/ssd indexes, between s3 and SQL. Gives us 60ms on a needle-in-a-haystack query.

0 0 0 51 0

View Details

Vinoth Chandar @byte_array

2 weeks ago

@Dipankartnt Every zero ETL basically adds query load and some CPU/network I/O to the source system. We are talking like we live in cartoon world where there is no physics :)