cozos edited a comment on issue #25899: [SPARK-29089][SQL] Parallelize blocking FileSystem calls in DataSource#checkAndGlobPathIfNecessary URL: https://github.com/apache/spark/pull/25899#issuecomment-546554524 With Parquet mode i.e. ``` spark.hadoop.fs.s3a.readahead.range 256K spark.hadoop.fs.s3a.input.fadvise random ``` ``` ______________________________________________________________________ | Type | fs.s3a.connection.maximum| Num Threads| Runtime(seconds)| |=====================================================================| | Flat Paths| 48 | 10 | 18.01 | | Flat Paths| 48 | 20 | 9.56 | | Flat Paths| 48 | 40 | 6.23 | | Flat Paths| 48 | 60 | 7.55 | | Flat Paths| 48 | 80 | 4.87 | | Flat Paths| 48 | 100 | 4.67 | | Flat Paths| 48 | 150 | 5.17 | | Flat Paths| 48 | 200 | 4.4 | | Flat Paths| 100 | 10 | 19.06 | | Flat Paths| 100 | 20 | 9.87 | | Flat Paths| 100 | 40 | 5.71 | | Flat Paths| 100 | 60 | 4.51 | | Flat Paths| 100 | 80 | 5.04 | | Flat Paths| 100 | 100 | 4.78 | | Flat Paths| 100 | 150 | 4.99 | | Flat Paths| 100 | 200 | 5.11 | | Flat Paths| 300 | 10 | 17.17 | | Flat Paths| 300 | 20 | 8.73 | | Flat Paths| 300 | 40 | 4.67 | | Flat Paths| 300 | 60 | 4.67 | | Flat Paths| 300 | 80 | 4.25 | | Flat Paths| 300 | 100 | 4.95 | | Flat Paths| 300 | 150 | 5.5 | | Flat Paths| 300 | 200 | 4.48 | | Glob Paths| 48 | 10 | 3.87 | | Glob Paths| 48 | 20 | 2.57 | | Glob Paths| 48 | 40 | 2.73 | | Glob Paths| 48 | 60 | 2.68 | | Glob Paths| 48 | 80 | 2.63 | | Glob Paths| 48 | 100 | 2.3 | | Glob Paths| 48 | 150 | 2.63 | | Glob Paths| 48 | 200 | 2.81 | | Glob Paths| 100 | 10 | 2.64 | | Glob Paths| 100 | 20 | 2.83 | | Glob Paths| 100 | 40 | 3.09 | | Glob Paths| 100 | 60 | 2.75 | | Glob Paths| 100 | 80 | 2.56 | | Glob Paths| 100 | 100 | 3.1 | | Glob Paths| 100 | 150 | 2.57 | | Glob Paths| 100 | 200 | 2.36 | | Glob Paths| 300 | 10 | 2.78 | | Glob Paths| 300 | 20 | 2.23 | | Glob Paths| 300 | 40 | 2.25 | | Glob Paths| 300 | 60 | 2.73 | | Glob Paths| 300 | 80 | 2.37 | | Glob Paths| 300 | 100 | 2.25 | | Glob Paths| 300 | 150 | 2.6 | | Glob Paths| 300 | 200 | 2.17 | FileSystem org.apache.hadoop.fs.s3a.S3AFileSystem: 0 bytes read, 0 bytes written, 151800 read ops, 0 large read ops, 0 write ops ```
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
