22 Introduction to RapidMiner Server Video Lecture | Rapid Miner: Learn and Master - Data & Analytics

Hello and welcome to this introduction to the RapidMiner Server.
In this tutorial we will go through the different sections of the web-interface of
the RapidMiner Server to give you a head-start for configuring your own platform.
During this video we will reference the module around, connecting and
executing processes on the server, frequently.
It is not absolutely necessary to familiarize yourself with that video prior to this one but it will be advantageous.
We will refer to it as the 'Server Basics' video.
Now the first question to answer when dealing with the server interface is: how to find it?
Well there are two ways. As we see, we have the RapidMiner Studio open here.
In the Repository panel you see the server repository which we established during the 'server basics'.
One very convenient way to access the server’s web interface is to just use the right mouse button
and select 'Administer (Web Interface)'.
Now the default browser is opening and we are presented with the login dialog.
And this represents the second option to get to the web interface - just open any browser and type:
http://yourservername:yourportnumber/faces/login.xhtml usually the port is ‘8080’
so try that in case you are not sure what yours is set to.
Alternatively you can also use the IP address instead of the domain-name.
We will login with the ‘admin’ account, which is one of the two special user accounts
together with the 'anonymous' user.
The admin user always does exist in any deployment and cannot be deleted
and there can always be only one with all administrator privileges.
The ‘anonymous’ user is for you to setup but once you create it then it is provided
with special rights automatically.
For more details check out the documentation under the entry here in the 'Documentation/Links' section or go to:
http://docs.rapidminer.com/ directly and search for 'Special Users'.
The first page-view we are presented with is the 'Home of the Repository' section.
Here we see a summary of recent events or recent places that have been visited.
From the ‘Browse repository’ section you can navigate through the folder structure.
If we navigate into the ‘data’ folder of the admin user's home-directory, we can see the ‘customer data’
which we copied and pasted there in the ‘server basics’ video.
If we select it we get a summary of where it initially was imported from and what it contains.
When we do the same on the process which we used in that same video - let's use the search page
and the term 'validation' to find it - we see that processes are actually stored as XMLs.
And as we expand it we can easily find the number of folds in our validation which we set to ‘1000’.
Here you can also "Create a cron trigger" if you want to run a process repeatedly or at a certain time.
We will do one for 7am each day.
A helpful public user interface to create the expression is for example http://www.cronmaker.com .
As soon as we click 'submit' - we are taken to the 'Schedules' view - which is in the 'processes' section.
We can see now our freshly scheduled process in this view.
Here you can pause the execution of scheduled jobs if needed.
The next entry in our menu is called 'executions'.
On this page we can see which processes are running and also the previous executions.
There is a filter and a search option provided in case you are looking for the status of a specific process.
In the table we see the past processes which have been run.
Those with a green tick mark have been completed successfully.
Should there be a little, red 'x' it means something has gone wrong.
You can then investigate that further through the log provided at the end of each entry.
Below the entry 'Executions' we can find 'Queues'.
Let's look at a default RapidMiner server installation briefly to help you understand what this is all about.
The big box represents a host or virtual machine on which the RapidMiner server is installed.
The orange box is the main application which has one execution option embedded
for executions related to web-services and web-apps.
The elements here represent a job container and a job agent respectively.
Together they form the second, execution option of RapidMiner Server.
Job agents can be seen as separate instances or applications on a machine
and they can hold one or more job containers.
Each Job agent is connected to a job queue within the RapidMiner server
and they can be installed on the same or alternatively a different host.
Job containers as the lowest hierarchical element are really execution applications
and can be somewhat compared to CPUs, but they are not restricted to using single CPUs.
In the default installation we have one job agent and one job container each.
Now on here you can see an example of a larger deployment with a total of four execution options.
One is, the execution engine within the server application, the second is the job container deployed on
the host-machine, and the third and fourth are job agents with containers,
which reside on external machines.
Two of the job agents have one job container each and the third has two, and each of the three job agents
is connected to a separate queue within the main server application.
You can also have multiple job agents connected to the same queue.
And as mentioned each agent can be connected to multiple, dedicated containers.
So you are really very flexible to scale this to become a very powerful computation platform.
Ok with this let's go back.
Through the link shown up here you will be taken to our docs page where you can find out
how to install additional job agents.
If you want additional job agents connect to another queue than the 'default' queue,
then you can and need to add one here.
If you need to change anything on the default queue setup - like alter who has access,
or add additional job agents to it - then you can do this here.
The next entry in this section provides the option to define web-services.
This page allows you to expose a RapidMiner process as a web-service.
You can invoke it via a URL and pass parameters in if needed.
We will just create a simple one here.
First thing is to give it a name. We will call it 'CrossValidation'.
If we click on the folder icon we can browse our server’s folders and select our
'05 cross validation on server' process.
Make sure to click 'select location'.
Choosing the right output format is very important as there are so many options.
We recently added OData format, which is open industry-standard used by vendors such as
SAP, Microsoft, Qlik and others to exchange data.
We leave the default here which is XML and will also leave the MIME type unchanged.
Just below you can add parameters to your web service.
To finish creating the service we hit 'submit' and now we can test it by clicking on the gear icon.
After it has been completed we get the performance vector returned, which is basically the content of the
‘confusion matrix’ as displayed in the results view of RapidMiner Studio.
In the 'server basics' tutorial we suggested to add a store operator in order to save your output,
but here is another way of retrieving the output information of processes.
On the right you get an iframe which you can copy and paste into your application or webpage
to invoke the web-service.
Alternatively you can use this direct link.
If we click that we get prompted for the user ID and the password - I will use the admin credentials –
the process is then executed and as output we get our performance vector again.
Now what else can we do in this section? We can define a trigger.
There is one which was already prepared so let's inspect it by clicking on the ‘notepad’ icon.
This one is set to monitor a file-change on the server.
So every time the file 'myfile.txt' changes the ‘cross validation’ process will be triggered.
Alternatively triggers can also listen to incoming mails.
We will not cover the remaining section shown here which is the administrative part of the server
but in essence here is where you define your security groups and users
and administer permission to features and folders.
Further you can define here your database connections and share them, your mail client settings
and the system settings of your server as well as licenses.
Last but not least on the lower left there is a link to the 'App designer'
which will allow you to create your dashboards.
Of course there are too many options around ‘App Designer’ to introduce them in this tutorial.
So we conclude our introduction to the server interface here.
During this module you have been introduced to the repository-browser, the process-scheduler,
you saw how to create a web service and how a trigger is set up and what it does.
Thank you very much for watching.