Set up Docker, Virtual Box, Cassandra in Docker - sharans003/DatabaseSystems GitHub Wiki

We'll be using Cassandra 2.1 and Ubuntu 16.04

Step 0: Setting up VirtualBox.

https://www.youtube.com/watch?v=lEvM-No4eQo

Also, to enable universal Copy Paste, Universal Drag n drop do the following:

  1. Devices -> Insert Guest Additions CD image (Keep saying YES or OK, let it do its thing)
  2. Devices -> Shared Clipboard -> Bidriectional
  3. Devices -> Drag and Drop -> Bidriectional

Step 1 : Installing Docker

https://docs.docker.com/install/linux/docker-ce/ubuntu/

Step 2: Installing Cassandra into docker

https://hub.docker.com/_/cassandra/ (Replace tag by 2.1)

  • sudo docker run --name sharan_cass --link some-cassandra:cassandra -d cassandra:2.1

Here is a snapshot of my terminal.

sudo apt-get remove docker docker-engine docker.io
sudo apt-get update
sudo apt-get install     apt-transport-https     ca-certificates     curl     software-properties-common
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
sudo apt-key fingerprint 0EBFCD88

sudo add-apt-repository    "deb [arch=amd64] https://download.docker.com/linux/ubuntu \
$(lsb_release -cs) \
stable"

sudo apt-get update
sudo apt-get install docker-ce
apt-cache madison docker-ce
pt-cache madison docker-ce
docker-ce | 17.12.0~ce-0~ubuntu | https://download.docker.com/linux/ubuntu xenial/stable amd64 Packages
apt-cache madison docker-ce
sudo apt-get install docker-ce=<VERSION>
sudo apt-get install docker-ce
sudo docker run hello-world
sudo groupadd docker
sudo usermod -aG docker $USER
docker run hello-world
sudo docker run hello-world
sudo chown "$USER":"$USER" /home/"$USER"/.docker -R
docker
sudo docker ps
docker ps
docker run --name sharan-cass -d cassandra:2.1
docker exec -it  sharan-cass  bash

Some quick docker commands that we need:

docker ps
docker ls
docker exec -it <container-name> bash
docker run --name cass1 -d cassandra:2.1

Step 3: Just to check if Casandra is working in the docker container, execute the below

  • cqlsh
  • exit To exit docker, type exit again

Step 4: Setting up DockerOps center

https://github.com/abh1nav/docker-cassandra

  • Skip everything till "Manual Mode"
docker pull abh1nav/cassandra:latest

Create a new container using opscenter
docker run -d --name opscenter -p 8888:8888 -p 61620:61620 abh1nav/opscenter:latest

Changing configuration of address.yaml in all the cassandra nodes that are part of a single cluster

Fields to be changed
--rpc_address- change from 0.0.0.0 to node ip address
--seed_ip- currently it will have ip of just the current node, add ip of all the other nodes in the cluster.
1. ssh into node:  docker exec -it NODENAME bash
2. Go to cd /etc/cassandra
3. Modify cassandra.yaml file


Install datastax agent on every node:
https://docs.datastax.com/en/archived/opscenter/5.1/opsc/install/opsc-agentInstallManual_t.html
Steps to be followed:
1. GO to cd /home
2. Run following command
curl -L http://downloads.datastax.com/community/datastax-agent-5.1.0.tar.gz | tar xz
3. cd datastax-agent-version
4. echo "stomp_interface: <<OPS_CENTER_IP>>" >> ./conf/address.yaml
5.bin/datastax-agent

Go to a directory
Pre-requisite to install agents:

apt-get update && apt-get install sudo
sudo apt-get install curl

Quick commands:

  1. Getting IP address:
OPS_IP=$(docker inspect -f '{{ .NetworkSettings.IPAddress }}' opscenter)

Optional Step: Installing Kitematic

  1. Go here: https://github.com/docker/kitematic/releases
  2. Download the .deb
  3. Install it.
  4. If that does not work try
sudo apt-get install kitematic
  1. Sometimes it might not show up as a logo. So in that case go to terminal and type kitemetic.

Step 5:

  1. Start new cassandra node which will be used as NTP Server.
  2. Make sure there are few ntp servers in /etc/ntpd.conf file.
  3. Start few nodes in a cluster.
  4. Install ntp on all nodes ( If you are using docker cassandra iamge, ntp should be pre installed).
  5. Setup ip address of ntp server node in /etc/ntpd.conf file on all the nodes. -- $pool.ntp.org maps to about 1000 low-stratum NTP servers. Your server will $pick a different set every time it starts up. Please consider joining the $pool: http://www.pool.ntp.org/join.html. server 172.17.0.4 iburst
  6. Restart ntp server on all the nodes using : /etc/init.d/ntp restart
  7. Try pinging ntp server from all the nodes and check on ntp server node if we are able to receive the ping. To ping : ping 172.17.0.4 . To check if node is receiving ping: tcpdump ip proto \icmp

Imp Links

https://askubuntu.com/questions/429306/ntpdate-no-server-suitable-for-synchronization-found https://github.com/cturra/docker-ntp

Topology 2:

  • We are going to have 3 VMS.
  • VM 1 is going to have the NTP server
  • VM 2 is going to have the 1st Cassandra node
  • VM 3 is going to have the 2nd Cassandra node

Setting broadcast IP:

  • In the first cassandra node, set the broadcast IP.
  • Do IFCONFIG in the cassandra node and get the IP.
docker run --name <container_name> -d -e CASSANDRA_BROADCAST_ADDRESS=192.168.89.4 -p 9044:9044  cassandra:2.1 docker container start <same_container_name> docker exec -it <same_container_name> bash
  • The broadcast IP is going to tell the other nodes in cassandra which IP to listen to.
  • In a way it is the "central" link for the rest of the cluster

Setting up the Seed IP:

  • The Seed IP is going to be set in the Cassandra node that is not the broadcast IP.
docker run --name <container_name> -d -e CASSANDRA_SEEDS=<broadcast_ip_that we set in node 1>

docker run --name <container_name> -d cassandra:2.1 -e CASSANDRA_SEEDS=<broadcast_ip_that we set in node 1>

Step : Connect Cassandra cluster across 2 machines, across 2 VMS

First step: Make sure both the VMS are on the same NAT

Second step: create a single node with just the broadcast address:

docker run --name <name> -d -e CASSANDRA_BROADCAST_ADDRESS=<ip of this vm> -p 7000:7000 cassandra 2.1

Third step: in the other VM

docker run --name <name> -d -e CASSANDRA_BROADCAST_ADDRESS=<ip of this vm> -e CASSANDRA_SEEDS=<broadcast mentioned in the first vm i.e. the first vm's IP address>-p 7000:7000 cassandra:2.1

Fourth step:

nodetool status

Working Second Topology:

  • Setup NTP

  • Say server ip is 10.10.0.1.

  • 1 Node is connected to server (0.debian.pool.ntp.org). Where ever we have pool, replace by server. Remove pool0,pool1 etc. Have just 1 line.

  • Also in fall back make it 127.0.0.1

  • Now in the clients, make the server IP point to 10.10.0.1, and remove the rest. Make the fall back 127.0.0.1

  • Now restart the server on all.

  • Then check if the nodes have synced.

  • If they have, move on with life.

  • Next Install Java. Make sure it is Java 1.8

  • java -version should do it

  • If its not 1.8 find out how to set it to that.

  • Now install cassandra from the documentation.

  • Check java version. It has to be 1.8, without that it wont work

  • Run cassandra.

Multi node cassandra: Server: 10.0.0.1 Client1: 10.0.0.2 Client2: 10.0.0.3

Server

In the yaml file do not change the server file at all

Client1

Go to client1: broadcast IP: 10.0.0.2 RPC IP: 10.0.0.2 Listen IP: 10.0.0.2 SeedIP: 10.0.0.2

Client2

Go to client2: broadcast IP: 10.0.0.3 RPC IP: 10.0.0.3 Listen IP: 10.0.0.3 SeedIP: 10.0.0.2, 10.0.0.3

Try the cassandra stress:

cassandra-stress write n=50000 -rate threads=40 -node d0.cass.org
⚠️ **GitHub.com Fallback** ⚠️