Database Systems Project - Comparative analysis of MongoDB, Apache HBase and Cockroach DB using YCSB benchmarking techniques

Installation Guide

CockroachDB Installation

For Linux based system

1. Download Binary

   curl https://binaries.cockroachdb.com/cockroach-v21.2.2.linux-amd64.tgz | tar -xz && sudo cp -i cockroach-v21.2.2.linux-amd64/cockroach /usr/local/bin/
   

2. Create directory for storing external libraries and copy library file

   mkdir -p /usr/local/lib/cockroach
   cp -i cockroach-v21.2.2.linux-amd64/lib/libgeos.so /usr/local/lib/cockroach/
   cp -i cockroach-v21.2.2.linux-amd64/lib/libgeos_c.so /usr/local/lib/cockroach/
   

   Note: Use sudo prefix for resolving permission errors.

3. Check Installation by running demo session

   cockroach demo
   

   Execute the query

   > SELECT ST_IsValid(ST_MakePoint(1,2));
   

   Result should be like

     st_isvalid
   +-------------+
       true
   (1 row)
   

For operating system other than Linux, refer to the link for installation.

Apache HBase Installation

1. Create a standard storage account on Azure with the following fields

   Region: Australia - East
   Performance : Standard
   Replication : Locally-redundant storage(LRS)
   

2. Create an HDInsight HBase cluster

   Region: Australia - East
   Cluster Type : HBase 2.1.6 (HDI 4.0)
   Details for the storage tab :
   Selection Method: Choose Radio button Use access key
   Storage account name: Enter the name of the Standard storage account created earlier
   Access Key:Enter the key1 access key for the standard storage account
   Head Node - E4 V3 (2),
   Zookeeper Node - A2 V2 (3),
   Region Node - D12 V2 (4)
   

Reference : https://techcommunity.microsoft.com/t5/analytics-on-azure-blog/azure-hdinsight-hbase-cluster-performance-comparison-using-ycsb/ba-p/1548530

MongoDB Installation

Part 1 : In each VM of the cluster (we are using a 3 VM cluster), install MongoDB with the following steps.

1. Update local packages.

   sudo apt-get update
   

2. Import MongoDB 5.0 public key used by package manager. This should return OK.

   wget -qO - https://www.mongodb.org/static/pgp/server-5.0.asc | sudo apt-key add -
   

3. Create MongoDB List file.

   echo "deb [ arch=amd64,arm64 ] https://repo.mongodb.org/apt/ubuntu focal/mongodb-org/5.0 multiverse" | sudo tee /etc/apt/sources.list.d/mongodb-org-5.0.list	
   

4. Install MongoDB packages.

   sudo apt-get install -y mongodb-org
   

Part 2 : Replica Set setup

1. Start the replica set daemon in each of the VMs.

   mongod --replSet "rs0" --bind_ip localhost,<Virtual NW IP of the machine>
   

2. Open another terminal while mongo daemon is running on the instances. Open mongo shell. This should be the primary node.

   mongo
   

3. Instantiate the replica set. Make sure all host ips are virtual network IPs.

   rs.initiate( {
      _id : "rs0",
      members: [
         { _id: 0, host: "<hostip1>:27017" },
         { _id: 1, host: "<hostip2>:27017" },
         { _id: 2, host: "<hostip3>:27017" }
      ]
   })
   

Success! You have now setup the 3 node replica set.

Installation Reference : https://docs.mongodb.com/manual/tutorial/install-mongodb-on-ubuntu/ Cluster - Replica set setup Reference : https://docs.mongodb.com/manual/tutorial/deploy-replica-set/

Running Instructions

CockroachDB

• In Azure Cloud:
  1. Create a separate VM instance in the same virtual network as cluster. This machine will be used as client to connect to cluster.
  2. Install Cockroach on the client and initialze the connector.

     cockroach init --insecure --host=<address of any node on --join list>
     

  3. Clone this repositiory on the machine.
  4. Run the following commands:

     cd scripts
     bash run_benchmark_cockroach.sh <Workload Class in upper case> <Hostname(IP) of Node 1> <Hostname(IP) of Node 2> <Hostname(IP) of Node 3>
     

  For eg. bash run_benchmark_cockroach.sh A 10.1.0.5 10.1.0.6 10.1.0.7

Apache HBase

• In Azure Cloud:
  1. Login to HDInsights cluster

     ssh <user-name>@<cluster-name>-ssh.azurehdinsight.net
     

  2. Launch HBase Shell and run the subsequent operations.

     hbase shell
     
      hbase_shell> n_splits = 100	
      hbase_shell> create 'usertable', 'cf', {SPLITS => (1..n_splits).map {|i| "user#{1000+i*(9999-1000)/n_splits}"}}
      hbase_shell> exit
     

  3. Clone this repository at the remote server

     cd scripts
     bash run_benchmark_hbase.sh <Workload Class in lower case> 
     

  For eg.
  bash run_benchmark_hbase.sh a

MongoDB

1. Start all the 3 instances of the cluster and SSH into them.

2. Start the mongod processes in each machine.

   sudo mongod --replSet rs0 --bind_ip localhost,<Machine IP> --dbpath mongoDBPath
   

3. Open a new terminal in the primary machine and run the benchmark with this command.

   bash run_benchmark_mongodb.sh a
   

Reproducing Figures

• Azure
  After executing the benchmarks, the output logs from each benchmark test are stored under the benchmark_output_logs directory. Then you can execute the following to produce images in benchmark_plots directory:

   cd scripts
   bash generate_figures_from_logs.sh