George Kobar
1.5K posts

George Kobar
@GeorgeKobar
Husband | Father | Elastic | Community Advocate | Lego Enthusiast| He/Him
Arvada, Colorado Katılım Ağustos 2012
366 Takip Edilen373 Takipçiler

@doris_apache You should look into Elasticsearch Query Language or ES|QL. It is a piped query language that delivers ease of use and similar functionality to SQL elastic.co/docs/explore-a…
English
George Kobar retweetledi

Spring AI 1.0 is out — congrats!
and yes, of course you can use it with elasticsearch:
spring.io/blog/2025/05/2…. I'll add a longer thread tonight but you can already take this as the preview
thanks to @starbuxman for the collaboration!


Spring AI@SpringAICentral
We’re excited to announce the general availability of Spring AI 1.0 GA! 🎉 A big thank you to everyone who contributed to making this release possible. 📢 Check out the release blog: spring.io/blog/2025/05/2…
English

Elasticsearch BBQ vs. OpenSearch FAISS: Vector search performance comparison elastic.co/search-labs/bl…
English

Elasticsearch redesigned from the ground up, seperating indexing from search that scales all in a fully managed solution.
Elastic@elastic
Do less with Serverless — Now GA 🎉 The fastest way to start and scale security, observability, and search hassle-free all powered by the same Elasticsearch you love, reimagined for the cloud. Get started now → go.es.io/3CVo82X
English
George Kobar retweetledi

Here is a talk I gave on Better Binary Quantization (BBQ) for vector search in Elasticsearch. For those that don't like reading blogs, or just want to know how bad the puns can really get, this talk is for you: youtube.com/watch?v=04NzMt…

YouTube
English
George Kobar retweetledi

🔥 As I said, we have been cooking! 🔥 Product Quantization is just too slow. Better Binary Quantization is faster than PQ is every metric with the same or better recall. elastic.co/search-labs/bl…
English

@kevins8 @elastic @OpenSearchProj If you look at other data, it paints an entirely different picture:
#overview" target="_blank" rel="nofollow noopener">ossinsight.io/analyze/elasti…
English

the google trends of @elastic vs @OpenSearchProj over the past 5 years is also interesting. elastic has fallen from its heights and stayed flat whereas opensearch is set to surpass it (blue is opensearch)

English

the @elastic announcement is going to be the first big case study of an "open source company" that switches its license and then changes its mind. i don't think they'll get back the community they lost to @OpenSearchProj but this might stop the bleeding
elastic.co/blog/elasticse…
English

George Kobar retweetledi

"Being able to call Elasticsearch and Kibana Open Source again is pure joy." — Shay Banon, Elastic Founder and CTO. Read more from @kimchy: go.es.io/4dNtVVR
#Elasticsearch
English

Elasticsearch is open source, again elastic.co/blog/elasticse… via @elastic
#elasticsearch #opensource
English
George Kobar retweetledi

the @StackOverflow developer survey of 2024 is out: survey.stackoverflow.co/2024/
specifically looking at the databases (blue is 2024, red is 2022):
1. don't bet against #postgresql
2. glad so see #elasticsearch holding up well while many others declined


English

How Orca leverages Search AI to help users gain visibility, achieve compliance, and prioritize risks elastic.co/blog/orca-sear… via @elastic #elasticsearch
English

@theoliveranwar Can you suggest a different exercise than leg extensions? I have been told these are very hard on your knees in particular ACL:
jospt.org/doi/pdf/10.251…
English

@GeorgeKobar @wabzqem Do you honestly think that the TLDR I’ve screenshoted is accurate and truthfully? «12x faster and therefore uses less computational resources»
English

Elastic busy misleading the community again.
Another benchmark where they use all 32 cpu cores to achieve lower latency. Then claim that they use fewer computational resources.
This is getting ridiculous. They only report 99p latency which is obviously lower when you use concurrent segment search instead of one thread 🤣

English

@jobergum I think you missed the point in the blog where we tested 3 items:
Elasticsearch ootb: ES out-of-the-box, with concurrent segment search;
OpenSearch ootb: without concurrent segment search enabled;
OpenSearch css: with concurrent segment search enabled
Example:

English

You compare latency of a single threaded sequential segment search with the latency of a parallelized search which will use all available CPU resources. If you want to compare latency, or the efficiency of two systems you need to include the throughout and utilization. Not just drop 12x faster without that context
English






