1
0
Fork 1
mirror of https://github.com/NixOS/nixpkgs.git synced 2024-12-25 03:17:13 +00:00
nixpkgs/pkgs/applications/networking/cluster/spark/default.nix
Kamil Chmielewski 3cdb8f581d spark: provided-hadoop -> nixpkgs hadoop
spark should use nixpkgs provided hadoop to make it consistent
for bigger Nix hadoop + spark setup
2018-04-23 20:48:33 +02:00

67 lines
2.2 KiB
Nix

{ stdenv, fetchzip, makeWrapper, jre, pythonPackages, coreutils, hadoop
, RSupport? true, R
, mesosSupport ? true, mesos
, version
}:
let
sha256 = {
"1.6.3" = "142hw73wf20d846l83ydx0yg7qj5qxywm4h7qrhwnd7lsy2sbnjf";
"2.2.1" = "10nxsf9a6hj1263sxv0cbdqxdb8mb4cl6iqq32ljq9ydvk32s99c";
}.${version};
in
with stdenv.lib;
stdenv.mkDerivation rec {
name = "spark-${version}";
src = fetchzip {
inherit sha256;
url = "mirror://apache/spark/${name}/${name}-bin-without-hadoop.tgz";
};
buildInputs = [ makeWrapper jre pythonPackages.python pythonPackages.numpy ]
++ optional RSupport R
++ optional mesosSupport mesos;
untarDir = "${name}-bin-without-hadoop";
installPhase = ''
mkdir -p $out/{lib/${untarDir}/conf,bin,/share/java}
mv * $out/lib/${untarDir}
sed -e 's/INFO, console/WARN, console/' < \
$out/lib/${untarDir}/conf/log4j.properties.template > \
$out/lib/${untarDir}/conf/log4j.properties
cat > $out/lib/${untarDir}/conf/spark-env.sh <<- EOF
export JAVA_HOME="${jre}"
export SPARK_HOME="$out/lib/${untarDir}"
export SPARK_DIST_CLASSPATH=$(${hadoop}/bin/hadoop classpath)
export PYSPARK_PYTHON="${pythonPackages.python}/bin/${pythonPackages.python.executable}"
export PYTHONPATH="\$PYTHONPATH:$PYTHONPATH"
${optionalString RSupport
''export SPARKR_R_SHELL="${R}/bin/R"
export PATH=$PATH:"${R}/bin/R"''}
${optionalString mesosSupport
''export MESOS_NATIVE_LIBRARY="$MESOS_NATIVE_LIBRARY"''}
EOF
for n in $(find $out/lib/${untarDir}/bin -type f ! -name "*.*"); do
makeWrapper "$n" "$out/bin/$(basename $n)"
substituteInPlace "$n" --replace dirname ${coreutils.out}/bin/dirname
done
ln -s $out/lib/${untarDir}/lib/spark-assembly-*.jar $out/share/java
'';
meta = {
description = "Apache Spark is a fast and general engine for large-scale data processing";
homepage = "http://spark.apache.org";
license = stdenv.lib.licenses.asl20;
platforms = stdenv.lib.platforms.all;
maintainers = with maintainers; [ thoughtpolice offline kamilchm ];
repositories.git = git://git.apache.org/spark.git;
};
}