buremba

34 ms·

391.
by buremba 9y ago
Would love to hear about the solution, are you going to implement it or publish an article?
392.
by buremba 9y ago
Do you guys know any way to calculate the intersection of two HLL instances? The algorithm allows merging two sets but it's not that easy to take the intersection of them. A + B - (A U B) is not the optimal way so I would love to hear
393.
by buremba 10y ago
Mainly the development cost. We trust OCR and have good experience with it, it's stable, fast and compact. We don't have any experience with Parquet and it's just too hard/risky to make this kind of conversion on a produ
394.
by buremba 10y ago
We have 300tb of orc data and can't convert it to Parquet. Also, ORC performs better for our use-case.
395.
by buremba 10y ago
ORC support?
396.
by buremba 10y ago
My point was they could convert their web applications to Chrome Apps which has better native api support and experience if Chrome Apps could become a thing.
397.
by buremba 10y ago
Yep, I hate that all of these companies try to convert their applications from tabs to dedicated processes. It would be a huge deal breaker if Chrome Apps could become a thing and replace them. Now I have 4 different Chrome apps running on
398.
by buremba 10y ago
Great, thanks!
399.
by buremba 10y ago
I will be waiting for you to release your clustering solution. Will it be also included in the open-source version or are you planning to release it as part of your commercial service?
400.
by buremba 10y ago
AFAIK, cstore_fdw (their columnar storage implementation) is not widely adopted among Citus users because it's not real-time. They also rely on memory to be able to query the data efficiently. I believe that the tricky part for time-s
401.
by buremba 10y ago
"..swagger repeats itself a lot with redundant annotations, it also leads to having more annotations than actual code." (Shameless plug) For this exact problem, we developed a Java library that makes easy to create RESTFul service
402.
by buremba 10y ago
I think it's not just "mental overhead". The main benefit would be able to use PipelineDB with other Postgresql extensions like CitusDB. It will make Postgresql competitor to many Hadoop oriented architectures.
403.
by buremba 10y ago
It looks like once you entered an invalid project name, the form doesn't accept a valid name. I fixed the issue and updated the app, thanks for reporting the bug. Nice humor BTW! :)
404.
by buremba 10y ago
It makes sense. I didn't about that API, thanks!
405.
by buremba 10y ago
The long term solution is usually building an in-house analytics service but it's not easy and cheap. We're also developing a generic analytics platform that can be installed with one-click to your favorite cloud provider so that
406.
by buremba 10y ago
Data management is a hard problem because most of the companies have different requirements and culture, that's why many "data-driven" companies try to build their in-house own analytics solution. In order to address this pr
407.
by buremba 10y ago
So you have a batching mechanism for creating TrailDB files, right? Then you're able to query historical metrics by processing these TrailDB files.
408.
by buremba 10y ago
Does TrailDB pre-aggregate metrics? AFAIK it stores the raw data bucketed with an actor id (usually a server or visitor) and performs compression and some columnar storage optimizations for events of actors.
409.
by buremba 10y ago
I don't think TrailDB is comparable to Beringei. TrailDB is built for behavioral analytics queries while Beringei is developed for time-series aggregated metrics.
410.
by buremba 10y ago
Because their product which is based on my base product will have a different name and branding. It won't be affiliated with my product so I'm not sure how it leads to faster ROI. The good thing is that my product will be based by
411.
Ask HN: Different investors on a GPL project, and conflict in IP?
2 points by buremba 10y ago|3 comments
412.
by buremba 10y ago
How does Scuba optimize the stored data-sets for aggregation queries compared to Presto (Raptor connector)? They both use common columnar data storage techniques such as compression, delta encoding and dictionary encoding. The main differen
413.
by buremba 10y ago
Shameless plug: I'm the maintainer of Rakam project. You don't have to deal with all of these complexities. There are open-source projects that can setup that infrastructure for you. For example, we provide Cloudformation scripts
414.
by buremba 10y ago
How does Scuba differ from Presto which is also developed by Facebook? It seems that it stores data in-memory and have data expiration feature but also has many common features such as SQL and distributed processing.
415.
by buremba 10y ago
However Peloton also aims to be an autonomous system. That's a lot for undergrad and grad students so I'm not sure if he wants Peloton to be stable in a near future.
416.
by buremba 10y ago
The lectures are great.
417.
by buremba 10y ago
I wonder why they try to support both OLTP and OLAP workloads. Supporting both of these workloads requires too much work (both row and columnar storage types, different algorithms for both storage and querying etc) and they didn't even
418.
by buremba 10y ago
It would be nice if they can share some info about Github::KV.
419.
SQL is still superior for big-data analytics(blog.rakam.io)
3 points by buremba 10y ago|0 comments
420.
by buremba 10y ago
How it handles that traffic? It would be great if they could share the architecture because using Mysql for analytics is not easy, you can't expect Mysql to handle concurrent queries that needs to aggregate billions of records when I v

More ›