2

I want get online data and save to cassandra keyspace. I read this guide, https://phoenixnap.com/kb/install-cassandra-on-windows, to run cassandra. It seems easy, but I get error that related to jdk. So, I tried different way. I tried with docker-toolbox (windows 8.1). I run cassndra by these steps in docker-toolbox shell:

$ docker run --name some-cassandra2 --network some-network -d cassandra:latest

$ docker run -it --network some-network --rm cassandra cqlsh some-cassandra2

enter image description here

cqlsh>create keyspace CityInfo with replication = {'class' : 'SimpleStrategy', 'replication_factor':2};

cqlsh>use CityInfo;

cqlsh> CREATE TABLE cities (id int,name text,country text,PRIMARY KEY(id));

cqlsh> CREATE TABLE users (username text,name text,age int,PRIMARY KEY(username));

Now, I want get online data and save to cities and users table with python code. I get online data. I try to connect with this code:

from cassandra.cluster import Cluster
cluster = Cluster(['172.18.0.2'],port=9042)
session = cluster.connect('cityinfo',wait_for_all_pools=False)
session.execute('USE cityinfo')
rows = session.execute('SELECT * FROM users')
for row in rows:
        print(row.age,row.name,row.username)

But I see error:

File "cassandra\cluster.py", line 3533, in cassandra.cluster.ControlConnection._reconnect_internal

NoHostAvailable: ('Unable to connect to any servers', {'172.18.0.2:9042': OSError(None, "Tried connecting to [('172.18.0.2', 9042)]. Last error: timed out")})

I tried several way. For example, I tried with other ip sush as 127.0.0.1:9042, or I added -p7000:7000 when runing cassandra to connect container port to device port. but I can't. Please guide me. What is the problem?

Ali Hosein pour
  • 220
  • 4
  • 19
  • Try using the container name instead of the ip – Hisham Jul 18 '21 at 07:17
  • In python code? I Tried. Doesn't work. – Ali Hosein pour Jul 18 '21 at 08:06
  • Where is the client code running? The container-private IP addresses are unreachable in almost all environments. From another container, it needs to be on the same Docker network, and the `cqlsh` invocation in the question is right. From outside Docker, on Docker Toolbox, you need the VM's IP address. – David Maze Jul 18 '21 at 12:54
  • I I worked all day on it. First, I wanted run cassandra on docker and python code in spyder. But as you mentioned, it's impossible. Then, I read different sources and tested them. One of them has been explained at this https://towardsdatascience.com/getting-started-with-apache-cassandra-and-python-81e00ccf17c9 . In this address, python code run on docker shell. I tried it step by step as mentioned, with two method. First I runed python code as container. It released same error. Second, I runed python code directly. It can't import cassandra-driver. Is there any solution? – Ali Hosein pour Jul 18 '21 at 15:06

1 Answers1

3

I suggest running your Python code from a container running on the same network, thus you can use the container name instead of the IP address directly in Python. I was able to run your code with no problem doing the following.

I created a docker container running Python, also running on some-network.

docker run -it --rm --network some-network python:3.8-slim bash

Proceeded to install cassandra-driver inside the container.

pip install cassandra-driver

Filled some dummy data in the users table and proceeded to open a Python terminal.

from cassandra.cluster import Cluster
cluster = Cluster(['some-cassandra2'], port=9042)
session = cluster.connect('cityinfo',wait_for_all_pools=False)
session.execute('USE cityinfo')
rows = session.execute('SELECT * FROM users')
for row in rows:
    print(row.age,row.name,row.username)

Notice that the only difference with your code is this line where I use the container name instead of the IP address:

cluster = Cluster(['some-cassandra2'], port=9042)
P. Cabaleiro
  • 66
  • 2