Giter VIP home page Giter VIP logo

Comments (1)

gruuya avatar gruuya commented on May 30, 2024

Finally got around to debugging this issue. It turns out this is a pathological case, since the file above has only one row group with 46 million rows:

$ du -h /Users/markogrujic/Downloads/area1.parquet
2.4G	/Users/markogrujic/Downloads/area1.parquet
$ parquet meta /Users/markogrujic/Downloads/area1.parquet

File path:  /Users/markogrujic/Downloads/area1.parquet
Created by: parquet-cpp version 1.5.1-SNAPSHOT
...


Row group 0:  count: 46234315  54.89 B records  start: 4  total(compressed): 2.363 GB total(uncompressed):2.363 GB
--------------------------------------------------------------------------------
         type      encodings count     avg size   nulls   min / max
X        DOUBLE    S RBR_    46234315  7.67 B     0       "-268.5730217805162" / "-68.00000035424296"
Y        DOUBLE    S RBR_    46234315  7.56 B     0       "-298.7632813729358" / "-108.00000066228672"
Z        DOUBLE    S RBR_    46234315  7.11 B     0       "-3.1149840059754865" / "97.65842918119495"
X_noise  DOUBLE    S RBR_    46234315  7.81 B     0       "-268.5741613914974" / "-67.99562912963299"
Y_noise  DOUBLE    S RBR_    46234315  7.68 B     0       "-298.7601570277598" / "-107.99524445315944"
Z_noise  DOUBLE    S RBR_    46234315  8.01 B     0       "-3.119372332345796" / "97.66003008163513"
R        DOUBLE    S RBR_    46234315  0.72 B     0       "-0.0" / "255.0"
G        DOUBLE    S RBR_    46234315  0.72 B     0       "-0.0" / "255.0"
B        DOUBLE    S RBR_    46234315  0.71 B     0       "-0.0" / "255.0"
time     DOUBLE    S RBR_    46234315  6.92 B     0       "1087.0830847224747" / "2118.595793055997"
eol      DOUBLE    S RBR_    46234315  0.00 B     0       "-0.0" / "1.0"
label    DOUBLE    S RBR_    46234315  0.00 B     0       "-0.0" / "8.0"

On the other hand, the arrow's async parquet reader ultimately streams the record batches after loading the whole row group in memory:

https://github.com/apache/arrow-rs/blob/c6184389241a0c85823aa494e8b5d93343771666/parquet/src/arrow/async_reader/mod.rs#L421-L423

It means that in this case the reader effectively loads the entire file in memory upon initiating the stream, which explains the memory profile.

Since there's nothing we can do in this case I'm closing this issue.

from seafowl.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.