This repository uses a simple idea of generating a tiny url, to create a scalable system on AWS cloud.
Uses python backend, and Cassandra database, with the python service hosted in AWS EC2 instances behind a load balancer, and Cassandra installed on a cluster of 2 AWS EC2 instances.
Source : https://phoenixnap.com/kb/install-cassandra-on-ubuntu
conda activate python3.8
Had some issues getting cassandra to work on WSL. To use cqlsh you'll need to use
cqlsh localhost --cqlversion="3.4.5"
You can start cassandra CLI using the command -
cqlsh
but for some weird reason it works only on python3.7, so set environment python3.7 for using cassandra cli
create keyspace bigcassandra with replication = {'class':'SimpleStrategy', 'replication_factor':3};
use bigcassandra;
List all tables in bigcassandra..
describe tables;
alter keyspace bigcassandra with replication = {'class':'SimpleStrategy', replication_factor:1} and Durable_writes=false;
keyspace can be thought of a database in SQL parlence
create table emp(emp_id int PRIMARY KEY, emp_name text, emp_city text, emp_sal varint, emp_phone varint);
Insert into table Insert into emp(emp_id, emp_name, emp_city, emp_sal, emp_phone) Values(1, 'test', 'test_city', 500000, 12345);
Create table UrlMap(tiny_url text PRIMARY KEY, url text);
Insert into UrlMap(tiny_url, url) Values ('http://mytinyurl.com/abcd', 'https://www.google.com')
curl -i -X POST -H "Content-Type: application/json" -d '{"url":"www.test.com"}' http://localhost:5000/add
Ensure the name of the cassandra table in Python model is all small letters - Phewww!!! resulted in a lot of time in debugging
In that case start the VM in the bridged networking mode. This causes an independent ip address to be assigned to this. You can use that IP address to access this service
This service is containerized into a docker container..
Using a multi-staged docker container build to cut down on the docker build runtime..
FROM python:3.8-slim as base_image
WORKDIR /app
COPY requirements.txt ./
RUN pip install --no-cache-dir -r requirements.txt
FROM base_image AS app_build
COPY . ./
CMD [ "python", "app.py" ]
Building the container
sudo docker build --target app_build -t system_design:tiny_url .
Ensure docker is running in WSL. The docker desktop service on Windows causes connectivity issues with Cassandra.
sudo service start docker
Start the container
sudo docker run --network=host -p 5000:5000 -it 0296699de4d4 /bin/bash
If the application is being run in the local environment, you'll need to set the environment as development before starting the application in docker shell
export FLASK_CONFIG="development"
python app.py
(https://stackoverflow.com/questions/54876879/connecting-cassandra-container-using-another-container)
Check if you can access a service from docker..
Run
nc -l 9999 in the base machine
Try to access the service from docker using
curl 127.0.0.1:9999 .... This will not work, as you cannot access the base machine from docker using the 127 addressing.
Check the ip of the docker using
ip addr show docker0
The ip shown there something like 172.17.0.1, is the ip of the machine that docker can see. Now try accessing the base machine using this IP.
curl 172.17.0.1:9999 ... This will work. Now change the cassandra connection ip from 127.0.0.1 to 172.17.0.1
Show the ip of the docker..
docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' <container_id>
Testing the API
pytest tests/test_api.py -s
To run some stress tests on the api's use the following command. This checks how long it takes to insert 10000 URLs and retrieve them. Test on this system shows..
(python3.7) pankaj@LAPTOP-1LSIP1HC:~/tiny_url$ pytest tests/test_load_api.py -s
======================================================================== test session starts ========================================================================
platform linux -- Python 3.8.5, pytest-6.1.1, py-1.9.0, pluggy-0.13.1
rootdir: /home/pankaj/tiny_url
collected 1 item
100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 10000/10000 [01:08<00:00, 145.62it/s]
100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 10000/10000 [00:37<00:00, 270.21it/s]
Inserting and retrieving 10000 url took 105.68551993370056 seconds
.
=================================================================== 1 passed in 105.80s (0:01:45) ===================================================================
pytest tests/test_load_api.py
Install cassandra as per the instructions on the above link on an Ubuntu EC2 instance.
Edit /etc/cassandra/cassandra.yaml with the following configurations..
Replace localhost with the private ip of the ec2 instance (in this case 172.31.30.180)
seeds will have the public ip of the ec2
seed_provider: # Addresses of hosts that are deemed contact points. # Cassandra nodes use this list of hosts to find each other and learn # the topology of the ring. You must change this if you are running # multiple nodes! - class_name: org.apache.cassandra.locator.SimpleSeedProvider parameters: # seeds is actually a comma-delimited list of addresses. # Ex: "<ip1>,<ip2>,<ip3>" - seeds: "52.66.206.122" #- seeds: "127.0.0.1:7000" rpc_address: 172.31.30.180 #private ip of ec2 listen_address: 172.31.30.180 #private ip of ec2 broadcast_address: 52.66.206.122 # public ip of ec2 instance endpoint_snitch: SimpleSnitch
Restart cassandra sudo systemctl restart cassandra
Connect to the cassandra instance
cqlsh 172.31.30.180 and create a keyspace create keyspace bigcassandra with replication = {'class':'SimpleStrategy', 'replication_factor':3};
Create a table
Create table UrlMap(tiny_url text PRIMARY KEY, url text);
Run the test case in your local pytest tests/test_api.py -s