binary tarball release asked.

[clojurians-org]

i plan to add datahub to nix package.

1. user can simple do the following to install datahub to local, and serval ingestion client tool to path.

   nix-env -iA linkedinDatahub
   cat mce.dat | datahub-mce-producer....

2. wrap datahub package with nix service module. so use can simple configure and launched it by [home-manager switch].

   {
   services.linkedin-datahub-gms = {
   enable = true ;
   package = pkgs.linkedDatahub ;
   kafkas = [ "PLAINTEXT://localhost:9092" ] ;
   neo4j = "" ;
   elasticsearch = "" ;
   postgresql = "" ;
   } ;
   services.linkedin-datahub-frontend = {
   enable = true ;
   package = pkgs.linkedDatahub ;
   gms ="localhost:..." ;
   } ;
   services.linkedin-datahub-pipeline = {
   enable = true ;
   package = pkgs.linkedDatahub ;
   kafka ="localhost:..."
   } ;
   }

the things i need is:

  src = fetchurl {
    url = "mirror://apache/kafka/${kafkaVersion}/kafka_${version}.tgz";
    inherit sha256;
  };

a related kafka package and service module examle is:

https://github.com/NixOS/nixpkgs/blob/master/pkgs/servers/apache-kafka/default.nix https://github.com/NixOS/nixpkgs/blob/master/nixos/modules/services/misc/apache-kafka.nix

usage:

  environment.systemPackages = [
    pkgs.apacheKafka
    pkgs.confluent-platform
  ]
  services.apache-kafka = {
    enable = true ;
    package = pkgs.apacheKafka ;
    logDirs = [ "/opt/nix-module/data/kafka" ] ;
    zookeeper = "localhost:2181" ;
    extraProperties = ''
      offsets.topic.replication.factor = 1
      zookeeper.session.timeout.ms = 600000
    '' ;    
  } ;
  services.confluent-schema-registry = {
    enable = true ;
    package = pkgs.confluent-platform ;
    kafkas = [ "PLAINTEXT://localhost:9092" ] ;
  } ;

[clojurians-org]

seperate server and client tools is better.

[mars-lan]

@clojurians-org do you plan to work on this further? If not we'll close this issue for now and maybe revisit it again in the future if needed.