Coder Social home page Coder Social logo

accumulo-docker's Introduction

Accumulo container, needs Hadoop HDFS and Zookeeper.

Simplest scenario, single node Hadoop, single node Zookeeper, single node Accumulo, no persistence. Accumulo uses hostname 'hadoop' for HDFS by default, and 'zookeeper' for Zookeeper.

  # Start Hadoop
  docker run -d --name hadoop cybermaggedon/hadoop:2.10.0

  # Start Zookeeper
  docker run -d --name zookeeper cybermaggedon/zookeeper:3.6.1

  # Start Accumulo, linking to other containers.
  docker run -d --name accumulo -p 9995:9995 -p 9997:9997 -p 9999:9999 \
        --link hadoop:hadoop \
	--link zookeeper:zookeeper \
        cybermaggedon/accumulo:1.9.3a

Default Accumulo root password is accumulo. To check it works:

  % docker ps | grep accumulo
  % docker exec -it insert-container-ID bash
  # /usr/local/accumulo/bin/accumulo shell -u root -p accumulo
  root@accumulo> tables

... and you should see a list of tables.

If you want to persist Accumulo you will need to ensure Hadoop and Zookeeper are persistent. Here we are using /data/hadoop and /data/zookeeper as persistent volumes.

e.g.

  # Start Hadoop
  docker run -d --name hadoop -v /data/hadoop:/data cybermaggedon/hadoop:2.10.0

  # Start Zookeeper
  docker run -d --name zookeeper -v /data/zookeeper:/data \
        cybermaggedon/zookeeper:3.6.1

  # Start Accumulo, linking to other containers.
  docker run -d --name accumulo  -p 9995:9995 -p 9997:9997 -p 9999:9999 \
        --link hadoop:hadoop \
        --link zookeeper:zookeeper \
	cybermaggedon/accumulo:1.9.3a

To do anything more complicated, we need to be able to manage the names or addresses that are provided to containers. Easiest way is to set up a user-defined network and allocate the IP addresses manually.

  ############################################################################
  # Create network
  ############################################################################
  docker network create --driver=bridge --subnet=10.10.0.0/16 my_network

  ############################################################################
  # HDFS
  ############################################################################

  # Namenode
  docker run -d --ip=10.10.6.3 --net my_network \
      --name=hadoop01 \
      -p 50070:50070 -p 50075:50075 -p 50090:50090 -p 9000:9000 \
      cybermaggedon/hadoop:2.10.0 /start-namenode

  # Datanodes
  docker run -d --ip=10.10.6.4 --net my_network --link hadoop01:hadoop01 \
      -e NAMENODE_URI=hdfs://hadoop01:9000 \
      --name=hadoop02 \
      cybermaggedon/hadoop:2.10.0 /start-datanode

  docker run -d --ip=10.10.6.5 --net my_network --link hadoop01:hadoop01 \
      -e NAMENODE_URI=hdfs://hadoop01:9000 \
      --name=hadoop03 \
      cybermaggedon/hadoop:2.10.0 /start-datanode

  ############################################################################
  # Zookeeper cluster, 3 nodes.
  ############################################################################
  docker run -d --ip=10.10.5.10 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=1 \
      --name zk1 -p 2181:2181 cybermaggedon/zookeeper:3.6.1
      
  docker run -d --ip=10.10.5.11 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=2 --name zk2 --link zk1:zk1 \
      cybermaggedon/zookeeper:3.6.1
      
  docker run -d --ip=10.10.5.12 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=3 --name zk3 --link zk1:zk1 \
      cybermaggedon/zookeeper:3.6.1

  ############################################################################
  # Accumulo, 3 nodes
  ############################################################################
  docker run -d --ip=10.10.10.10 --net my_network \
      -p 50095:50095 -p 9995:9995 \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.10 \
      -e GC_HOSTS=10.10.10.10 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e SLAVE_HOSTS=10.10.10.10,10.10.10.11,10.10.10.12 \
      -e MONITOR_HOSTS=10.10.10.10 \
      -e TRACER_HOSTS=10.10.10.10 \
      --link hadoop01:hadoop01 \
      --name acc01 cybermaggedon/accumulo:1.9.3a

  docker run -d --ip=10.10.10.11 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.11 \
      -e GC_HOSTS=10.10.10.10 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e SLAVE_HOSTS=10.10.10.10,10.10.10.11,10.10.10.12 \
      -e MONITOR_HOSTS=10.10.10.10 \
      -e TRACER_HOSTS=10.10.10.10 \
      --link hadoop01:hadoop01 \
      --name acc02 cybermaggedon/accumulo:1.9.3a

  docker run -d --ip=10.10.10.12 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.12 \
      -e GC_HOSTS=10.10.10.10 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e SLAVE_HOSTS=10.10.10.10,10.10.10.11,10.10.10.12 \
      -e MONITOR_HOSTS=10.10.10.10 \
      -e TRACER_HOSTS=10.10.10.10 \
      --link hadoop01:hadoop01 \
      --name acc03 cybermaggedon/accumulo:1.9.3a

If you want persistence, Hadoop and Zookeeper need a volume. Accumulo uses HDFS for state, so no volumes needed.

  ############################################################################
  # Create network
  ############################################################################
  docker network create --driver=bridge --subnet=10.10.0.0/16 my_network

  ############################################################################
  # HDFS
  ############################################################################

  # Namenode
  docker run -d --ip=10.10.6.3 --net my_network \
      --name=hadoop01 \
      -p 50070:50070 -p 50075:50075 -p 50090:50090 -p 9000:9000 \
      -v /data/hadoop01:/data \
      cybermaggedon/hadoop:2.10.0 /start-namenode

  # Datanodes
  docker run -d --ip=10.10.6.4 --net my_network --link hadoop01:hadoop01 \
      -e NAMENODE_URI=hdfs://hadoop01:9000 \
      --name=hadoop02 \
      -v /data/hadoop02:/data \
      cybermaggedon/hadoop:2.10.0 /start-datanode

  docker run -d --ip=10.10.6.5 --net my_network --link hadoop01:hadoop01 \
      -e NAMENODE_URI=hdfs://hadoop01:9000 \
      --name=hadoop03 \
      -v /data/hadoop03:/data \
      cybermaggedon/hadoop:2.10.0 /start-datanode

  ############################################################################
  # Zookeeper cluster, 3 nodes.
  ############################################################################
  docker run -d --ip=10.10.5.10 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=1 \
      -v /data/zk1:/data \
      --name zk1 -p 2181:2181 cybermaggedon/zookeeper:3.6.1
      
  docker run -d --ip=10.10.5.11 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=2 --name zk2 --link zk1:zk1 \
      -v /data/zk2:/data \
      cybermaggedon/zookeeper:3.6.1
      
  docker run -d --ip=10.10.5.12 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=3 --name zk3 --link zk1:zk1 \
      -v /data/zk3:/data \
      cybermaggedon/zookeeper:3.6.1

  ############################################################################
  # Accumulo, 3 nodes
  ############################################################################

  # First node is master, monitor, gc, tracer, tablet server
  docker run -d --ip=10.10.10.10 --net my_network \
      -p 50095:50095 -p 9995:9995 \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.10 \
      -e GC_HOSTS=10.10.10.10 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e SLAVE_HOSTS=10.10.10.10,10.10.10.11,10.10.10.12 \
      -e MONITOR_HOSTS=10.10.10.10 \
      -e TRACER_HOSTS=10.10.10.10 \
      --link hadoop01:hadoop01 \
      --name acc01 cybermaggedon/accumulo:1.9.3a

  # Two slave nodes, tablet server only
  docker run -d --ip=10.10.10.11 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.11 \
      -e GC_HOSTS=10.10.10.10 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e SLAVE_HOSTS=10.10.10.10,10.10.10.11,10.10.10.12 \
      -e MONITOR_HOSTS=10.10.10.10 \
      -e TRACER_HOSTS=10.10.10.10 \
      --link hadoop01:hadoop01 \
      --name acc02 cybermaggedon/accumulo:1.9.3a

  docker run -d --ip=10.10.10.12 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.12 \
      -e GC_HOSTS=10.10.10.10 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e SLAVE_HOSTS=10.10.10.10,10.10.10.11,10.10.10.12 \
      -e MONITOR_HOSTS=10.10.10.10 \
      -e TRACER_HOSTS=10.10.10.10 \
      --link hadoop01:hadoop01 \
      --name acc03 cybermaggedon/accumulo:1.9.3a

Now, the above configuration has a problem, in that it uses the default Accumulo startup. This runs all of the Accumulo processes in the background inside each container. If Accumulo processes fail, there's no restart. The Docker Way to do this is to run each Accumulo process in its own container, so that when the process fails, the container fails, which can be detected and restarted. To do this, I invoke /start-process and provide the name of an Accumulo process to stat.


  ############################################################################
  # Create network
  ############################################################################
  docker network create --driver=bridge --subnet=10.10.0.0/16 my_network

  ############################################################################
  # HDFS
  ############################################################################

  # Namenode
  docker run -d --ip=10.10.6.3 --net my_network \
      --name=hadoop01 \
      -p 50070:50070 -p 50075:50075 -p 50090:50090 -p 9000:9000 \
      -v /data/hadoop01:/data \
      cybermaggedon/hadoop:2.10.0 /start-namenode

  # Datanodes
  docker run -d --ip=10.10.6.4 --net my_network --link hadoop01:hadoop01 \
      -e NAMENODE_URI=hdfs://hadoop01:9000 \
      --name=hadoop02 \
      -v /data/hadoop02:/data \
      cybermaggedon/hadoop:2.10.0 /start-datanode

  docker run -d --ip=10.10.6.5 --net my_network --link hadoop01:hadoop01 \
      -e NAMENODE_URI=hdfs://hadoop01:9000 \
      --name=hadoop03 \
      -v /data/hadoop03:/data \
      cybermaggedon/hadoop:2.10.0 /start-datanode

  ############################################################################
  # Zookeeper cluster, 3 nodes.
  ############################################################################
  docker run -d --ip=10.10.5.10 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=1 \
      -v /data/zk1:/data \
      --name zk1 -p 2181:2181 cybermaggedon/zookeeper:3.6.1
      
  docker run -d --ip=10.10.5.11 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=2 --name zk2 --link zk1:zk1 \
      -v /data/zk2:/data \
      cybermaggedon/zookeeper:3.6.1
      
  docker run -d --ip=10.10.5.12 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e ZOOKEEPER_MYID=3 --name zk3 --link zk1:zk1 \
      -v /data/zk3:/data \
      cybermaggedon/zookeeper:3.6.1

  ############################################################################
  # Accumulo, 3 nodes
  ############################################################################

  # Master
  docker run -d --ip=10.10.10.10 --net my_network \
      -p 50095:50095 \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.10 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e GC_HOSTS=10.10.10.11 \
      -e SLAVE_HOSTS=10.10.10.12,10.10.10.13,10.10.10.14 \
      -e MONITOR_HOSTS=10.10.10.15 \
      -e TRACER_HOSTS=10.10.10.16 \
      --link hadoop01:hadoop01 \
      --name acc-master cybermaggedon/accumulo:1.9.3a /start-process master

  # GC
  docker run -d --ip=10.10.10.11 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.11 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e GC_HOSTS=10.10.10.11 \
      -e SLAVE_HOSTS=10.10.10.12,10.10.10.13,10.10.10.14 \
      -e MONITOR_HOSTS=10.10.10.15 \
      -e TRACER_HOSTS=10.10.10.16 \
      --link hadoop01:hadoop01 \
      --name acc-gc cybermaggedon/accumulo:1.9.3a /start-process gc

  # Slave 1
  docker run -d --ip=10.10.10.12 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.12 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e GC_HOSTS=10.10.10.11 \
      -e SLAVE_HOSTS=10.10.10.12,10.10.10.13,10.10.10.14 \
      -e MONITOR_HOSTS=10.10.10.15 \
      -e TRACER_HOSTS=10.10.10.16 \
      --link hadoop01:hadoop01 \
      --name acc-slave1 cybermaggedon/accumulo:1.9.3a /start-process tserver

  # Slave 2
  docker run -d --ip=10.10.10.13 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.13 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e GC_HOSTS=10.10.10.11 \
      -e SLAVE_HOSTS=10.10.10.12,10.10.10.13,10.10.10.14 \
      -e MONITOR_HOSTS=10.10.10.15 \
      -e TRACER_HOSTS=10.10.10.16 \
      --link hadoop01:hadoop01 \
      --name acc-slave2 cybermaggedon/accumulo:1.9.3a /start-process tserver

  # Slave 3
  docker run -d --ip=10.10.10.14 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.14 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e GC_HOSTS=10.10.10.11 \
      -e SLAVE_HOSTS=10.10.10.12,10.10.10.13,10.10.10.14 \
      -e MONITOR_HOSTS=10.10.10.15 \
      -e TRACER_HOSTS=10.10.10.16 \
      --link hadoop01:hadoop01 \
      --name acc-slave3 cybermaggedon/accumulo:1.9.3a /start-process tserver

  # Monitor - this has the web server.
  docker run -d --ip=10.10.10.15 --net my_network \
      -p 9995:9995 \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.15 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e GC_HOSTS=10.10.10.11 \
      -e SLAVE_HOSTS=10.10.10.12,10.10.10.13,10.10.10.14 \
      -e MONITOR_HOSTS=10.10.10.15 \
      -e TRACER_HOSTS=10.10.10.16 \
      --link hadoop01:hadoop01 \
      --name acc-monitor cybermaggedon/accumulo:1.9.3a /start-process monitor

  docker run -d --ip=10.10.10.16 --net my_network \
      -e ZOOKEEPERS=10.10.5.10,10.10.5.11,10.10.5.12 \
      -e HDFS_VOLUMES=hdfs://hadoop01:9000/accumulo \
      -e NAMENODE_URI=hdfs://hadoop01:9000/ \
      -e MY_HOSTNAME=10.10.10.16 \
      -e MASTER_HOSTS=10.10.10.10 \
      -e GC_HOSTS=10.10.10.11 \
      -e SLAVE_HOSTS=10.10.10.12,10.10.10.13,10.10.10.14 \
      -e MONITOR_HOSTS=10.10.10.15 \
      -e TRACER_HOSTS=10.10.10.16 \
      --link hadoop01:hadoop01 \
      --name acc-tracer cybermaggedon/accumulo:1.9.3a /start-process tracer

If volumes don't mount because of selinux, this command may be your friend:

chcon -Rt svirt_sandbox_file_t /path/of/volume

The following environment variables are used to tailor the Accumulo deployment:

  • ZOOKEEPERS: A comma separated list of hostnames or IP addresses of the Zookeeper servers. Defaults to zookeeper, useful for a single stand-alone Zookeeper.
  • HDFS_VOLUMES: A comma separated list of HDFS URIs for volumes to store Accumulo data. I've only tested with one volume. Defaults to hdfs://hadoop:9000/accumulo.
  • NAMENODE_URI: The URI of the Hadoop name-node. Defaults to hdfs://hadoop:9000/accumulo.
  • INSTANCE_NAME: A unique name for this Accumulo instance. Defaults to accumulo.
  • MY_HOSTNAME: Hostname or IP address to share with other nodes in the cluster. Defaults to localhost, useful only for a standalone cluster.
  • GC_HOSTS, MASTER_HOSTS, SLAVE_HOSTS, MONITOR_HOSTS and TRACER_HOSTS list of hostnames or IP addresses of nodes which run these daemons. Defaults to something useful for a single-node cluster.

The following environment variables are used to tailor Accumulo sizing. The default values are very low, suitable for a development environment, but nothing more significant.

  • MEMORY_MAPS_MAX: Default 8M.
  • CACHE_DATA_SIZE: Default 2M.
  • CACHE_INDEX_SIZE: Default 2M.
  • SORT_BUFFER_SIZE: Default 5M.
  • WALOG_MAX_SIZE: Default 5M.

The values Accumulo developers recommend are:

  • 512MB:
    • MEMORY_MAPS_MAX: 80M
    • CACHE_DATA_SIZE: 7M
    • CACHE_INDEX_SIZE: 20M
    • SORT_BUFFER_SIZE: 50M
    • WALOG_MAX_SIZE: 100M
  • 1GB:
    • MEMORY_MAPS_MAX: 256M
    • CACHE_DATA_SIZE: 15M
    • CACHE_INDEX_SIZE: 40M
    • SORT_BUFFER_SIZE: 50M
    • WALOG_MAX_SIZE: 256M
  • 2GB:
    • MEMORY_MAPS_MAX: 512M
    • CACHE_DATA_SIZE: 30M
    • CACHE_INDEX_SIZE: 80M
    • SORT_BUFFER_SIZE: 50M
    • WALOG_MAX_SIZE: 512M.
  • 3GB:
    • MEMORY_MAPS_MAX: 1G
    • CACHE_DATA_SIZE: 128M
    • CACHE_INDEX_SIZE: 128M
    • SORT_BUFFER_SIZE: 200M
    • WALOG_MAX_SIZE: 1G

Source at http://github.com/cybermaggedon/accumulo-docker.

accumulo-docker's People

Contributors

cybermaggedon avatar

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.