Now that you have successfully installed Meltano locally or in the cloud, you're ready to start create your first project, connect data sources, build pipelines, perform analysis and create dashboards.
The following instructions assume you are able to access Meltano's user interface locally from
http://localhost:5000 or at
To initialize a new project, open your terminal and navigate to the directory that you'd like to store your Meltano projects in.
Next, to create your project, you will use the
meltano init command which takes a
PROJECT_NAME that is of your own choosing. For this guide, let's create a project called "carbon."
Meltano shares anonymous usage data with the team through Google Analytics. This is used to help us learn about how Meltano is being used to ensure that we are making Meltano even more useful to our users.
If you would prefer to use Meltano without sending the team this data, learn how to configure this through our environment variables docs.
meltano init carbon
This will create a new directory named
carbon in the current directory and initialize Meltano's basic directory structure inside it.
Inside the Meltano project directory, all plugin configuration (which may include tokens and passwords) is stored inside the
which is automatically added to the project's
.gitignore file to prevent this potentially sensitive information from accidentally being pushed up to a hosted Git repository.
Now that you've created your first Meltano project, let's change directory to our new project and start Meltano UI:
cd carbon meltano ui
Meltano is now running and should open a new tab at http://localhost:5000.
You should see now see the Extractors page, which contains various options for connecting your data source.
For this guide, we will be following the "Fast Path" to get you up and running as quickly as possible. So let's install
tap-carbon-intensity by clicking on the
Install button inside its card. While an Extractor is installed, you are presented with the following modal:
tap-carbon-intensity Extractor does not require any configuration (like a username or password). Once the installation is complete, you will progress to the next step: "Load".
Once you've setup your Extractor, you will be greeted with the Loaders page:
Now that Meltano is pulling data in from your data source(s), you need to choose where and in what format you would like that data stored.
target-postgres for this project by clicking
Install in its card.
While it is installing, make sure that your PostreSQL database is up and running. If you need help with this, check out our PostgreSQL database tutorial.
Once it is finished installing, you will see the following modal:
target-postgres is configured with a database named
warehouse that can be customized if desired. Once you configure your database as desired, click
With our extractor and loader configured, you should now be greeted with the Schedules page with a modal to create your first pipeline!
Meltano provides Orchestration using Apache Airflow, which allows you to create scheduled tasks to run pipelines automatically. For example, you may want a recurring task that updates the database at the end of every business day.
In the current form, you will see:
All we need to do is click
Save to start our new pipeline! The pipeline's log opens automatically and you can check the pipeline running and what Meltano does behind the scenes to extract and load the data. You should see a spinning icon that indicates that the pipeline is not completed:
Once it's complete, the indicator will disappear and you should be able to see the final results of the extraction:
You can click the
Analyze button to select a model to analyze. The same
Analyze button is available inline within your pipeline. If you close the log and go back to the Pipelines page you can check the log of any past pipeline by clicking the
Log button next to it:
Congratulations! Now that you have connected a data source, configured a target database, and run a successful pipeline for the dataset, we are now ready to analyze the data!
There are currently three ways to select a data model for analyzing, exploring, and report building:
Analyzedropdown (all data models)
Analyzebutton (contextual data models)
Analyzebutton (contextual data models)
Selecting a data model from one of these options takes us to the next step.
The Analyze page contains an interactive user interface to allow you to dynamically build queries and visualize your data. By default, it will run a standard report.
Now, you can explore and analyze the
tap-carbon-intensity data by clicking on different Attributes in the Query side menu on the left hand side. This will trigger live changes to the SQL queries which will update the chart dynamically.
While exploring the Analyze page, you can also check out:
When we find an analysis that we want to reference in the future, we can easily do this by creating a report. This can be accomplished by clicking on the
Save Report dropdown in the Analyze toolbar. This will open a dropdown with a default report name that is dynamically populated, but can be easily changed.
Once we click
Save, we should see the upper left "Untitled Report" change to our new report name.
And with that, our analysis has been saved!
As you acquire more reports, you will probably want to organize them via dashboards. This can be done by clicking on the new
Add to Dashboard dropdown in the toolbar.
Since we have never created a dashboard, click on
New Dashboard, which will trigger a modal that contains a dynamically generated dashboard name that can be customized as desired.
Once we click
Create, we can now verify that our report has been added to the Dashboard by clicking on the
Add to Dashboard menu.
We can also visit the Dashboard directly by clicking on the
Dashboard navigation item in the header, which shows our newly created Dashboard.
Once we select it, you should see a similar page to below:
And with that, you have now setup a complete end-to-end data solution with Meltano! 🎉
To learn about more Meltano recipes and functionality with our tutorials.