Skip to content

Remote files

In this tutorial, we learn how to load data from remote files into Qlik Associative Engine using the gRPC File Connector API. The tutorial uses the connector example from core-grpc-s3-file-connector.


To follow along in this tutorial, you should have basic understanding of Docker.

You need the following software installed:

  • Git
  • Docker
  • docker-compose
  • Node.js (optional)
  • corectl (optional)


Shell commands should be run in a Bash shell. If you are using Windows, we recommend using Git Bash.

The gRPC S3 File Connector

The S3 File Connector example is a connector built to display how to be able to download and upload files to and from an S3 bucket in AWS. The connector is built in Node.js and uses the gRPC protocol to stream data to and from the Qlik Associative Engine.


To run the example code, clone the qlik-oss/core-grpc-s3-file-connector Git repository. Before you continue, look at the documentation to get familiar with the content and structure of the repository.


In your shell, make sure current directory is at the repository root.

The connector example repo provides a docker-compose.yml, that starts the example connector together with a Qlik Associative Engine.

version: "3.3"

    image: qlikcore/engine:<version>
      - 9076:9076
    # EnableGrpcFileStreamConnector parameter enables the feature.
    # Default is disabled (0)
    # GrpcConnectorPlugins parameter configures the connector in engine.
    # Format <connector name>,<host>:<port>.
    # The "fs" at the end of the string marks that the connector
    # supports streaming of files
    command: -S EnableGrpcFileStreamConnector=1 -S GrpcConnectorPlugins="s3-grpc-file-connector,s3-grpc-file-connector:50051,fs" -S AcceptEULA=${ACCEPT_EULA}
    build: ../ # Build the connector from the Dockerfile
      - 50051:50051 # Expose the connector on port 50051
      # The following variables are passed as environment variables to the connector upon startup
      CORE_S3_FILE_CONNECTOR_BUCKET_NAME: ${CORE_S3_FILE_CONNECTOR_BUCKET_NAME} # Name of the S3 bucket that the connector should target

In this example we are going to use a data set from airports.csv. This data set is present in an S3 Bucket in AWS. The name of the S3 bucket is configured in the example connector by passing it as an environment variable; CORE_S3_FILE_CONNECTOR_BUCKET_NAME. Set the region of the S3 bucket in the environment variable: CORE_S3_FILE_CONNECTOR_BUCKET_REGION. List of regions can be found here: AWS Regions To be able to access the S3 bucket the connector will also need to provide the credentials, which in this example is the access key id CORE_S3_FILE_CONNECTOR_BUCKET_ACCESS_KEY_ID and a secret CORE_S3_FILE_CONNECTOR_BUCKET_SECRET_ACCESS_KEY.

To run this example you will need to set up your own S3 bucket and configure the name and credentials as described above. You can either hardcode the variables in the docker-compose.yml, or set them as environment variables. You will also need to copy the airports.csv file into your bucket, or switch to using another data set.

To start the example connector and a Qlik Associative Engine just run:

cd examples
ACCEPT_EULA=yes docker-compose up -d

Creating a connection

Once a connector has been properly configured in Qlik Associative Engine it will be possible to create and use the connector in a connection.

The connector that was configured in Qlik Associative Engine using the GrpcConnectorPlugins parameter, was configured with type s3-grpc-file-connector.

In a corectl.yml configuration file it would now be possible to create a connection object like this:

    type: s3-grpc-file-connector

This will create a connection with name s3bucket and type s3-grpc-file-connector. Further description on how connections are configured in corectl can be found here.

This tutorial also provides a Node.js example where a connection is created using enigma.js, how this is done can be seen here.


In the examples above the connection string is empty and not used. The connection can however also include a connection string, username and password which will also be passed to the connector. In many cases it will be up to the connector to define which information that should be passed by the Qlik Associative Engine.

Loading data from file in a S3 bucket

This section will describe how to load data from a file using the connection that was set up in the previous section. As mentioned earlier we have a data set available in the S3 bucket named airports.csv.

Imagine that we want to load all the data from that file using a load script. That can be done using this load script statement:

FROM [lib://s3bucket/airports.csv]
(txt, codepage is 28591, delimiter is ',', msq);

The lib statement is basically lib://<connection name>/<remote file>.

Storing data into a file in a S3 bucket

Similar to loading data from a file with this feature, it is also possible to store files to a remote share.

For a Qlik Core user this is usable for example when storing data to QVD files. Considering the table being loaded in previous section, we can now store data from the table Airports into a QVD file using the following load script syntax:

FROM Airports INTO [lib://s3bucket/exported.qvd];

This will store the data into a file named exported.qvd located in the S3 bucket.

Using file script functions in a load script

In addition to only being able to load from and store into a remote share, there are also a number of load script functions related to files that can be used.

Below is a subset of the file script functions that can be used together with this feature.

LOAD Distinct
  FileName() as filename,
  FileTime() as filetime,
  FileSize() as filesize
FROM [lib://s3bucket/];

The load script above will load all file names, sizes and last modified times of the files in the S3 bucket, and add it to a table called Files.

Additional file functions that can be used are further described in the script reference section for file functions.

Running the examples

The connector repo contains two different examples. One is written in Node.js using enigma.js and one is utilizing the tool corectl.

Both examples perform the same set of actions:

  1. Creates an app in Qlik Associative Engine
  2. Creates a connection in Qlik Associative Engine that utilizes the gRPC S3 File Connector
  3. Sets a load script
  4. Performs a reload

To run the Node.js example:

cd node
npm install
npm start

And to run the corectl example:

cd corectl
corectl build
corectl get tables

Next steps

More runnable load script examples are available in the qlik-oss/core-data-loading Git repository.