News for hackers. Stuff that matters.
Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by WhitneyLand on Wednesday January 25, 2017 @11:13AM (#13482589)
I don't see how these results provide much useful information in terms of being able to say x is faster than y.

The hardware doesn't seem consistent across different benchmarks. He says it's fast for a "cpu system", but for practical purposes Phi competes more with GPGPUs.

Would this be just as fast with one redis system with 512GB ram? I don't know too many apples to oranges here.

Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by sologub on Wednesday January 25, 2017 @11:27AM (#13482758)
The author isn't testing just the software but combinations of various software/hardware systems, including for example PostgreSQL on an i5 CPU, 16GB RAM and 850 SSD hardware: http://tech.marksblogg.com/billion-nyc-taxi-rides-postgresql...
Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by WhitneyLand on Wednesday January 25, 2017 @11:43AM (#13482950)
Great, and btw I think it's cool and I enjoyed reading it.

But what useful conclusions can be drawn from it?

Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by LeifCarrotson on Wednesday January 25, 2017 @12:23PM (#13483471)
All conclusions are only valid for similar workloads, but each of MapD and GPUs, Q/kdb+ and Xeon Phi, Redshift, Athena, Big Query, Presto, and Elasticsearch claim to be fast, inexpensive, easy to work with, and otherwise great for Big Data. Which ones really are fast? How fast is fast? How much is this going to cost? Do I need 5 nodes or 50?

A few examples of some useful conclusions:

- Just because a relatively well-optimized PostgreSQL database on a regular workstation takes 5 minutes to run a query doesn't mean you can't get special hardware to run that query faster than you can type.

- Spark + S3 + Amazon Elastic Map Reduce look like an ideal tool to work with large data, but they're pretty slow compared to better tools, and even compared to plain PostgreSQL.

- HDFS really is a lot faster than S3.

- Performance of an Xeon Phi 64-core CPU is within an order of magnitude to an NVidia Titan X.

- Loading 104 GB of compressed data into Q/kdb+ expands to 125 GB with and takes about 30 minutes, but on Redshift expands to 2 TB and takes many hours to upload on a normal connection, plus 4 hours to actually import!

- It might cost $5000 to custom-build a GPU-based supercomputer that can do these queries in under a second, but you can run similar queries if you're willing to wait for 5 minutes each by spinning up instances for a few dollars an hour plus a few more dollars an hour for storage, or by just running PostgreSQL on your workstation.

Also, not a conclusion, but it's incredibly useful to have a simple example exactly how to configure the tool and import some CSV data

Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by WhitneyLand on Wednesday January 25, 2017 @04:13PM (#13486168)
These conclusions don't seem very useful because either they are already well established or are not valid. Some examples:

Just because a relatively well-optimized PostgreSQL database on a regular workstation takes 5 minutes to run a query doesn't mean you can't get special hardware to run that query faster than you can type.

Already well established for years with systems like redis, and more recently with gpu databases, and other techniques posted on HN regularly.

Spark + S3 + Amazon Elastic Map Reduce...is pretty slow compared to better tools, and even compared to plain PostgreSQL.

Not valid because it doesn't generalize. It so much depends on type of work being done, system architecture, etc, that you can only say it may or may not be true.

HDFS really is a lot faster than S3.

This is already well established, Amazon states aa much right in the docs: http://docs.aws.amazon.com/emr/latest/ManagementGuide/emr-pl...

Performance of an Xeon Phi 64-core CPU is within an order of magnitude to an NVidia Titan X.

Not precise enough to matter because getting within 10x difference is not close to being competitive.*

Loading 104 GB of compressed data into Q/kdb+ expands to 125 GB with and takes about 30 minutes, but on Redshift expands to 2 TB and takes many hours to upload on a normal connection, plus 4 hours to actually import!

I don't see how it's possible for 104GB of csv text data to decompress into only 125GB. For cvs to compress only ~20%...doesn't make sense.

It might cost $5000 to custom-build a GPU-based supercomputer that can do these queries in under a second

No, two problems here. The hardware in question could have used 1 cheap CPU instead of two expensive Xeons and been much less expensive. Bigger problem: The MapD software itself will be $50,000.

Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by LeifCarrotson on Wednesday January 25, 2017 @05:09PM (#13486710)
The speed comparisons may be well known to you, but as someone only really using trivial desktop app SQLite databases, they weren't known to me. Thanks for pointing out my errors!

> I don't see how it's possible for 104GB of csv text data to decompress into only 125GB. For cvs to compress only ~20%...doesn't make sense.

The CSV file itself is around 500 GB. The internal representation, which might use binary formats for numbers, or compress text, uses 125 GB. Redshift expands it to 2TB for all the indexing and mapping.

> Bigger problem: The MapD software itself will be $50,000.

Ouch. That's a rather large oversight. Is the author affiliated with MapD, perhaps?

Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by juliangoldsmith on Wednesday January 25, 2017 @03:48PM (#13485911)
>Just because a relatively well-optimized PostgreSQL database on a regular workstation takes 5 minutes to run a query doesn't mean you can't get special hardware to run that query faster than you can type.

This is actually my biggest complaint with the article. He used cstore_fdw with Postgres, which doesn't allow much real indexing, and as far as I can tell (knowing only a little bit about it) he didn't really use any of the benefits of cstore_fdw.

I'd be interested to see how plain Postgres, possibly on a compressed filesystem, with properly-indexed tables stacks up.

Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by sologub on Wednesday January 25, 2017 @11:45AM (#13482982)
At least that Xeon Phi is a great piece of engineering! :)
Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by lorenzhs on Wednesday January 25, 2017 @11:24AM (#13482722)
A single machine doesn't have the required memory bandwidth to do this in the same time.
Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by WhitneyLand on Wednesday January 25, 2017 @12:01PM (#13483182)
Why not? Sounds like the initial data load and indexing are done up front. Once you get past that to run the benchmark it's not clear that a quad channel ddr4 system would be saturated.
Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by lorenzhs on Wednesday January 25, 2017 @04:56PM (#13486609)
A current-generation Xeon E7 has a memory bandwidth of up to 102 GB/s. The article says 90, which is probably a realistic estimate of achievable throughput. But the data is 500GB. So the problem is not with loading it from disk (that can be done beforehand), but getting it from CPU to RAM fast enough.
Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by throwawayish on Thursday January 26, 2017 @07:57AM (#13490929)
Inaccurate; the very point of columnar stores is that if I'm only interested in a column I only expend the memory bandwidth required for that very column and nothing else. Hence typical queries to a columnar store would never stream the whole 500 GB for processing.
Re:1.1B Taxi Rides on Kdb+/q and 4 Xeon Phi CPUs by lorenzhs on Thursday January 26, 2017 @12:15PM (#13493158)
Right, a columnar store would touch only a couple (1.1bn * three numerical columns) of those 500 GB. But the question was about Redis :)

The author has an overview of the benchmarks with various systems at http://tech.marksblogg.com/benchmarks.html – but one can't just compare rows as many factors vary (esp. hardware).