Have docker for mac installed and run your notebooks with following config:
SparkConf().setMaster('spark://master:7077')
Make sure your jobs have spark.executor.cores
to less than 4 and spark.executor.memory
to 512mb max.
An example of this is:
import os
config = {
"--deploy-mode" : "client",
"--supervise" : "",
"--num-executors" : "2",
"--executor-memory" : "512M",
"--driver-memory": "512M",
"--jars" : "/home/jovyan/work/deps/xxxx",
"--packages" : "com.crealytics:spark-excel_2.11:0.9.5;org.apache.hadoop:hadoop-aws:2.6.0,com.amazonaws:aws-java-sdk-s3:1.11.213,com.amazonaws:aws-java-sdk-core:1.11.213,com.databricks:spark-csv_2.11:1.3.0"
"pyspark-shell"
}
cmd = ""
for flag in config.keys():
val = config[flag]
cmd = ("%s %s %s "%(cmd, flag, val))
os.environ['PYSPARK_SUBMIT_ARGS'] = cmd
import pyspark
conf = pyspark.SparkConf().setAppName('ExcelLoader').setMaster('spark://master:7077')
conf.set("spark.executor.cores","2")
conf.set("spark.shuffle.service.enabled", "false")
conf.set("spark.dynamicAllocation.enabled", "false")
conf.set("spark.io.compression.codec", "snappy")
conf.set("spark.rdd.compress", "true")
conf.set("spark.executor.memory", "512mb")
sc = pyspark.SparkContext.getOrCreate(conf)
sc.setLogLevel('INFO')
from pyspark.sql import SQLContext
sql_context = SQLContext(sc)
hadoopConf = sc._jsc.hadoopConfiguration()
hadoopConf.set("fs.s3n.awsAccessKeyId", os.environ['AWS_ACCESS_KEY'])
hadoopConf.set("fs.s3n.awsSecretAccessKey", os.environ['AWS_SECRET_KEY'])
Please note that you will have to add all jars to a deps folder local to the docker-compose.yml file.
├── LICENSE
├── Makefile <- Makefile with commands like `make data` or `make train`
├── README.md <- The top-level README for developers using this project.
├── data
│ ├── external <- Data from third party sources.
│ ├── interim <- Intermediate data that has been transformed.
│ ├── processed <- The final, canonical data sets for modeling.
│ └── raw <- The original, immutable data dump.
│
├── deps <- Jar files you want to load to your spark cluster
│
├── docker-compose.yml <- 5 worker spark cluster
│
├── docs <- A default Sphinx project; see sphinx-doc.org for details
│
├── models <- Trained and serialized models, model predictions, or model summaries
│
├── notebooks <- Jupyter notebooks. Naming convention is a number (for ordering),
│ the creator's initials, and a short `-` delimited description, e.g.
│ `1.0-jqp-initial-data-exploration`.
│
├── references <- Data dictionaries, manuals, and all other explanatory materials.
│
├── reports <- Generated analysis as HTML, PDF, LaTeX, etc.
│ └── figures <- Generated graphics and figures to be used in reporting
│
├── requirements.txt <- The requirements file for reproducing the analysis environment, e.g.
│ generated with `pip freeze > requirements.txt`
│
├── src <- Source code for use in this project.
│ ├── __init__.py <- Makes src a Python module
│ │
│ ├── data <- Scripts to download or generate data
│ │ └── make_dataset.py
│ │
│ ├── features <- Scripts to turn raw data into features for modeling
│ │ └── build_features.py
│ │
│ ├── models <- Scripts to train models and then use trained models to make
│ │ │ predictions
│ │ ├── predict_model.py
│ │ └── train_model.py
│ │
│ └── visualization <- Scripts to create exploratory and results oriented visualizations
│ └── visualize.py
│
└── tox.ini <- tox file with settings for running tox; see tox.testrun.org
Project based on the cookiecutter data science project template. #cookiecutterdatascience