InformationSystems

ArangoDB

Ubuntu 20.04 -

https://www.arangodb.com/download-major/ubuntu/ (via package manager)

sudo apt update
sudo apt install curl apt-transport-https gnupg2 -y

First add the repository key to apt like this:

curl -OL https://download.arangodb.com/arangodb39/DEBIAN/Release.key
sudo apt-key add - < Release.key

Add the ubuntu repository

echo 'deb https://download.arangodb.com/arangodb39/DEBIAN/ /' | sudo tee /etc/apt/sources.list.d/arangodb.list

Refresh apt

sudo apt update

Install arangodb by running

sudo apt install arangodb3=3.9.2-1

Check status

sudo systemctl status arangodb3

Enable the WEB UI
for data-configuring the instance etc

vim /etc/arangodb3/arangod.conf

# Replace the line of endpoint = tcp.. , with
# endpoint = tcp://ip-server-of-server-or-domain:8529
# Save changes and restart the arangodb service

sudo systemctl restart arangodb3

For this to work we need to open port 8529 on our vm(vps provider)

Usage

ArangoDB can be used as a graph db, a document db or a key-value store db.
Can be used as an elastic search replacement(for search)
builtin-fox javascript framework for nodejs microservices (with great performance)
Good for geo-spacial searches also. Natively supports polygon searches.

docker run -e ARANGO_NO_AUTH=1 -p 8529:8529 -d --name test-arangodb arangodb

Visit localhost:8529
Create a collection named friends

insert { name: "Will" } into friends
// Cmd+Enter to run query
// Key is like a primary id
return document("friends/151")
return document(["friends/514", "friends/122"])

for friend in friends
    return friend

for friend in friends
    filter friend.name == "Will"
    // filter friend.age > 5
    return friend

// Join
for friend in friends
    for state in states
    filter friend.state == state._key
    // Return anything we like
    return {friend, state: state.name}

Use it as key-value

Create a collection names sessions

insert {_key: "abcde", logins: 1} into sessions

// Simply fetch it since we have specified the key
return document("sessions/abcde");

Usage with arangosh(shell client)

https://www.arangodb.com/docs/stable/programs-arangosh.html

python2.7 ./bin/ycsb load arangodb -s -P workloads/workloada -p arangodb.ip="localhost" -p arangodb.port=8529

Riak KV

Riak KV is a distributed key-value NoSQL database designed to deliver maximum data availability by distributing data across multiple servers. As long as your Riak KV client can reach one Riak server, it should be able to write data.

cd riak

docker-compose up -d coordinator

# navigate to http://localhost:8098/admin/
# https://docs.riak.com/riak/kv/latest/using/admin/riak-control/index.html
# read more https://hub.docker.com/r/basho/riak-kv/

Chose Riak® KV flexible key-value data model for web scale profile and session management, real-time big data, catalog, content management, customer 360, digital messaging, and more use cases.
Riak KV Tour

YCSB

Follow official instructions to install the Yahoo Cloud Serving Benchmark
https://github.com/brianfrankcooper/YCSB

cd ycsb/lib

# Add logging lib
curl -L -O 'https://search.maven.org/remotecontent?filepath=org/slf4j/slf4j-simple/1.7.25/slf4j-simple-1.7.25.jar'

Workloads info https://github.com/brianfrankcooper/YCSB/wiki/Core-Workloads

Properties brianfrankcooper/YCSB#1206

-target : the target number of operations per second. By default, the YCSB Client will try to do as many operations as it can. For example, if each operation takes 100 milliseconds on average, the Client will do about 10 operations per second per worker thread. However, you can throttle the target number of operations per second. For example, to generate a latency versus throughput curve, you can try different target throughputs, and measure the resulting latency for each.

Varagos/InformationSystems