By Mihalis Tsoukalos | Nov 07, 2019

Secure layer to insert data into Apache Kafka

Secure layer to insert data into Apache Kafka

Lenses provides a secure access layer to streaming data for any Apache Kafka (self-managed, AWS MSK, HDInsight etc.).

As well as being able to query data via SQL, you can also produce data into a topic. This adds an extra level of security and auditing since Lenses works at the application layer and is protected by fine-grained security via namespaces.

Let’s take a look at four different ways of injecting data into a Kafka Topic via Lenses.


To make it easier to follow this guide please download the free Lenses Docker “Box” image from here.

Additionally, you should get the Lenses CLI here.

As Lenses comes with sample Kafka topics populated with data, this tutorial will use an existing Kafka topic.

Using Lenses UI

There exist two distinct ways to insert data into a Kafka topic from Lenses UI. By inserting your own events or by inserting data that is automatically generated by Lenses.

Inserting your own data

From the SQL Studio page, simply write the following SQL code:

INSERT INTO cc_data(_key,number,customerFirstName,customerLastName,country,currency,blocked)
        VALUES ("Using-sql-studio","SQL","Studio","test","","",false);

One way of verifying data has been inserted is via the Topics page and selecting cc_data topic.

insert lenses

Alternatively, you can execute the following SELECT query in SQL Studio:

SELECT * FROM cc_data where _key='Using-sql-studio'

Using Auto Generated Data

From the Topics page, select cc_data topic.

From the bottom on upper-right side of page, select Insert Records.

auto lenses

By clicking Autogenerate button Lenses will create an event based on random data that conforms to the schema such as:

        "key": "...",
        "value": {
            "number": "icrxkbkvHpk",
            "customerFirstName": "TUMOeiPvUsLYdyJTKeBNp",
            "customerLastName": "sdkNpJQeKlIwWpJenkxEGN",
            "country": "fGgMjNFTbTuBmKGOpFyJRUTOj",
            "currency": "yntJh",
            "blocked": false

auto-data lenses

The key for the record can be either specified by you or by Lenses and Kafka. In this case it is specified by us and will be using-auto-generate. After that you will need to press the + Insert Records button and you are done.

The following SELECT query executed in SQL Studio will verify that the data is in cc_data:

SELECT * FROM cc_data where _key='using-auto-generate'

Using Lenses Data API

The remaining three ways will use Lenses Data API to insert data into an existing Kafka topic.

Using Lenses CLI

Execute the lenses-cli shell command that will put you in the interactive lenses-cli shell. The following output shows the interaction that allows you to insert data into a Kafka topic:

lenses-cli shell
    __                                 ________    ____
   / /   ___  ____  ________  _____   / ____/ /   /  _/
  / /   / _ \/ __ \/ ___/ _ \/ ___/  / /   / /    / /
 / /___/  __/ / / (__  )  __(__  )  / /___/ /____/ /
/_____/\___/_/ /_/____/\___/____/   \____/_____/___/
Docs at
Connected to [http://localhost:3030] as [admin], context [box]
Use "!" to set output options [!keys|!keysOnly|!stats|!meta|!pretty]
Crtl+D to exit
lenses-sql> INSERT INTO cc_data(_key,numlenses-sql> INSERT INTO
    VALUES ("Using lenses-cli","test","test","test","","",false);
{"flag":true,"info":"1 records inserted"}

You can verify that the record is there by executing the following SQL statement:

lenses-sql> SELECT * FROM cc_data WHERE _key='Using lenses-cli' LIMIT 1;

Using the Command Line

This method uses a command line utility like curl(1) to access Lenses API and insert data into a Kafka topic. You will need to have your data stored in JSON format.

In order to connect to Lenses using curl(1), you will need to have a authentication token that is connected to your current connection. In order to get that authentication token you will need to execute the following command first:

curl -H 'Content-Type: application/json' --request POST --data
'{"user":"admin","password":"admin"}' http://localhost:3030/api/login

The output of that command will be the authentication token, which will have a format similar to 16a8b2dd-b2d0-449e-8203-9dd51d7950cf. As this functionality is embedded in the presented bash(1), you will not need to remember the authentication token.

Lenses also provides the ability to create service account tokens.

The data will be in a JSON file named /tmp/myData.json with the following contents:

                "key": "using-command-line-bash",
                "value": {
                        "number": "123456789",
                        "customerFirstName": "Lenses",
                        "customerLastName": "Kafka HTTP",
                        "country": "GR",
                        "currency": "Euro",
                        "blocked": false

bash(1) shell script named will automate the process and make it more user friendly:

#!/usr/bin/env bash
if [ "$#" -lt 1 ]
    echo "Usage: $0 filename"
if [[ ! -e "${filename}" ]]
    echo "${filename} does not exist!"
# get token
token=$(curl -H 'Content-Type: application/json' --request POST --data \
'{"user":"admin","password":"admin"}' http://localhost:3030/api/login 2>/dev/null)
# send data
curl -H "X-Kafka-Lenses-Token:${token}" --request POST --data @${filename} ${endpoint} -w '\n\n' 2>/dev/null

The script accepts one command line parameter, which is the name the of the JSON file. Executing will create the following kind of output:

./ /tmp/myData.json
The request has been fulfilled and resulted in a new resource being created.

You can verify that the insert operation was successful using lenses-cli shell:

lenses-sql> SELECT * FROM cc_data WHERE _key='using-command-line-bash' LIMIT 1;
{"number":"123456789","customerFirstName":"Lenses","customerLastName":"Kafka HTTP","country":"GR","currency":"Euro","blocked":false}

Creating a Client in Python 3

Here we are going to develop a Lenses client. We will use Python 3 but you can use any programming language you want provided that it supports the needed functionality.

The name of the Python 3 utility is and contains the following code:

#!/usr/bin/env python3
import json
from requests import get, delete, post, put
from json import loads
from pprint import pprint as echo
import websocket
VALIDATE_SQL_QUERY = '/api/sql/validation'
SQL_END_POINT = "/api/ws/v2/sql/execute?"
class myPythonClient():
    def __init__(self, url, username, password):
        self.url = url
        self.username = username
        self.password = password
        self.test = self.connect()
        self.default_headers = {'Content-Type': 'application/json', 'Accept': 'application/json', 'x-kafka-lenses-token': self.token}
        self.default_headers_2 = {'Content-Type': 'text/event-stream', 'Accept': 'text/event-stream', 'x-kafka-lenses-token': self.token}
    def connect(self):
        LOGIN = "/api/login"
        login_url = self.url + LOGIN
        payload = {'user': self.username, 'password': self.password}
        default_headers = {'Content-Type': 'application/json', 'Accept': 'application/json, text/plain'}
        response = post(login_url, data=json.dumps(payload), headers=default_headers)
        if response.status_code != 200:
            raise Exception("Could not connect to the API [{}]. Status code [{}]. Reason [{}]"\
            .format(login_url, response.status_code, response.reason))
            self.token = response.text
            if self.token == None:
                raise Exception("Cannot recieve Token.")
        AUTH = "/api/auth"
        auth_url = self.url + AUTH
        new_headers = {"X-Kafka-Lenses-Token": self.token}
        response = get(auth_url, headers=new_headers)
        if response.status_code != 200:
            raise Exception("Could not connect to the API [{}]. Status code [{}]. Reason [{}]"\
             .format(auth_url, response.status_code, response.reason))
            self.token = response.json().get("token", None)
        return self.token
    def execSQL(self, query=None):
        self.query = query
        self.params = {'sql': self.query}
        if 'https' in self.url:
            url = self.url.replace("https", "wss")+SQL_END_POINT
            url = self.url.replace("http", "ws") + SQL_END_POINT
        ws = websocket.create_connection(url)
        message = {
            "token": self.token,
            "sql": self.query,
        data_list = []
        stats_list = []
        while True:
            temp_data = loads(ws.recv()) 
            temp_type = temp_data.get("type", None)
            if temp_type is None:
                raise KeyError("There isn't key 'type'")
            if temp_type == "RECORD":
            elif temp_type == "STATS":
            elif temp_type == "END":
if __name__ == "__main__":
    data = myPythonClient(url="http://localhost:3030",username="admin",password="admin")
    data.execSQL(query="""INSERT INTO cc_data(_key,number,customerFirstName,customerLastName,country,currency,blocked)\
    data.execSQL(query="""SELECT * FROM cc_data WHERE _key='using-python' LIMIT 1""")

Do not be put off by the code, it is used for illustrating the capabilities of Lenses and how easy it is to insert data to a Kafka topic.

The script has two main parts. The first one has to do with connecting to Lenses and uses the connect() function whereas the second one is about executing the desired SQL statement with the help of the execSQL() function.

Note that you might need to install some extra Python 3 package for to work on your local machine.

Executing will generate the following output:

[{'rownum': 0, 'value': {'flag': True, 'info': '1 records inserted'}}]
[{'key': 'using-python',
  'metadata': {'__keysize': 12,
               '__valsize': 25,
               'offset': 26,
               'partition': 0,
               'timestamp': 1572955657396},
  'rownum': 0,
  'value': {'blocked': False,
            'country': '',
            'currency': '',
            'customerFirstName': 'test',
            'customerLastName': 'test',
            'number': 'Python'}}]

The output of verifies that the INSERT statement was successful.

Additionally, the image that follows will also verify that the data is inserted in the desired Kafka topic.

python lenses


Using Lenses as a secure data access layer is a great way to safely open up Apache Kafka to more developers, users and applications. Learn more about Lenses from the different usecases or get access to a free trial.

Useful Links

Ready to get started with Lenses?

Try now for free