rxin

8 ms·

31.
by rxin 8y ago
Apache Spark doesn't use Calcite.
32.
by rxin 8y ago
This (天网, or "sky net") most likely comes from proverb "天网恢恢疏而不漏" (tian wang hui hui shu er bu lou), meaning "Justice has long arms." This saying was from Lao Zi, going back thousands of years.
33.
by rxin 9y ago
What's bloated about it? From the UX point of view, it's pretty lightweight for most common operations. Goes into the app, finds the contact, sends message, done. There is virtually no difference between WeChat, WhatsApp, and Appl
34.
by rxin 9y ago
I live in the Bay Area and I use both WhatsApp and WeChat and I would say WeChat is a significantly better product, with more features and better video/audio quality. When I visit China, WeChat is another order of magnitude more indisp
35.
Meltdown and Spectre’s Performance Impact on Big Data Workloads in the Cloud(databricks.com)
4 points by rxin 9y ago|0 comments
36.
by rxin 9y ago
WeChat is pretty big in Hong Kong now. I was on Hong Kong MTR last week and half of the people standing around me used WeChat within 10 mins.
37.
Introducing Vectorized UDFs for PySpark(databricks.com)
2 points by rxin 9y ago|0 comments
38.
by rxin 9y ago
Spark actually has probably the best standard SQL support among all open source big data frameworks. It can run all of TPC-DS queries without modifications.
39.
Developing Custom Machine Learning Algorithms in PySpark(databricks.com)
2 points by rxin 9y ago|0 comments
40.
by rxin 9y ago
What you said about Spark was not true. One can totally interactively query a Kafka stream using Spark, in SQL.
41.
Databricks Raises $140M from Top VCs in Mission to Bring AI to 'The 99%'(forbes.com)
15 points by rxin 9y ago|2 comments
42.
by rxin 9y ago
DJI AFAIK moved out of Hong Kong due to the lack of funding. Hong Kong never developed any serious engineering culture & talent base and it is very difficult to start a high tech venture in Hong Kong, from a funding point of view. From
43.
A Vision for Making Deep Learning Simple(databricks.com)
5 points by rxin 9y ago|0 comments
44.
Top 5 Reasons for Choosing S3 Over HDFS(databricks.com)
2 points by rxin 9y ago|0 comments
45.
Debugging a failing test case caused by query running “too fast”(databricks.com)
42 points by rxin 10y ago|7 comments
46.
Databricks and Apache Spark 2016 Year in Review(databricks.com)
1 points by rxin 10y ago|0 comments
47.
Scalable Partition Handling for Cloud-Native Architecture in Apache Spark 2.1(databricks.com)
1 points by rxin 10y ago|0 comments
48.
by rxin 10y ago
Take a look at the customers featured at reinvent. Computing as "utility" is the future, and the future is arriving fast.
49.
by rxin 10y ago
The one-off cost is very high, and there is higher ongoing maintenance cost as well. Most organizations are moving to the cloud, because it in general makes more economic sense.
50.
by rxin 10y ago
I'd argue this (performance/cost) is exactly the right metric to measure. One of the biggest benefits of the cloud is elasticity. In on-premise world, one would have to provision based on peak demand, and most of the time the clus
51.
by rxin 10y ago
I definitely agree that it'd be great to decouple the improvements in software and drops in cloud pricing. In reality it is pretty difficult because the Nanjing U/Alibaba team spent a lot of time also optimizing the software speci
52.
Setting a new world record in CloudSort with Apache Spark(databricks.com)
86 points by rxin 10y ago|21 comments
53.
by rxin 10y ago
There were tons of competition for wechat, but wechat crushed them due to a combination of superior product and stronger platform (with QQ to start with). As a matter of fact, usually startups in China (or large companies pushing for a new
54.
by rxin 10y ago
Reynold from Databricks and the Apache Spark project here. Would you mind shooting me an email rxin at databricks.com so I can understand more the issues you run into?
55.
by rxin 10y ago
Why did you say TPC-H sucks and TPS-DS is the thing to use? My impression is that while newer, TPC-DS is actually easier to optimize for.
56.
Voice from CERN: Spark 2.0 Performance Improvements Investigated with Flame Graphs(databricks.com)
1 points by rxin 10y ago|0 comments
57.
Introducing Apache Spark 2.0(databricks.com)
3 points by rxin 10y ago|0 comments
58.
Introducing Apache Spark 2.0(databricks.com)
55 points by rxin 10y ago|1 comments
59.
Databricks Community Edition: A Free Learning Platform for Apache Spark(databricks.com)
4 points by rxin 10y ago|0 comments
60.
Apache Spark 2.0 Preview: Machine Learning Model Persistence(databricks.com)
4 points by rxin 10y ago|0 comments

More ›