[Music] in this video we'll be covering and demonstrating the mid server mid stands for management instrumentation and discovery mid servers are also an important piece of the architecture when dealing with integrations the mid server is a java agent that runs on a windows or linux machine to allow us to securely communicate devices in our organization it does this through outbound https connections to our instance on port 443 so there's no need to create risky inbound port configurations on our firewall for larger enterprises we may want multiple mid servers for fault tolerance load balancing or to reduce network traffic between the mid server and affected systems for example if our main headquarters is in scotland and we've got a database server we want to interface with in italy it would be wise to place a mid server in the italy data center of course work with your network team to determine the best placement for your network configuration once we've determined where to place the mid server or servers validate that the system running the agent has port 443 access to the instance this is pretty standard stuff but if your server has any special security configuration it may be an issue see the documentation for full details on configuring the mid server system security now it's time to install the software we can do this manually or by going to guided setup itom guided setup and clicking the get started which is fine for getting our basic mid server up and running and right at the top we see the mid server section with three straightforward tasks so let's get started the first task is to create a mid-server user account this is a good idea in most integrations to have a specific account rather than a real user in case that user changes roles authentication or potentially terminated we don't want our mid server to stop running because someone was forced to change their password so we'll create one called mid.demo and give it a password that was easy if we like we can create additional mid server accounts but for now let's move on by marking this task complete next we download and install the software by following the next step on the guided setup we're taken to the download page this is also available from mid server downloads here we pick the proper download for our operating system for example we'll be installing to a linux host and have a couple manual steps to take the windows installation is simply a matter of following the installer instructions in the interest of time we won't go through all the details of installing and configuring the software suffice it to say there's a link in the embedded help on your instance to point you to the right direction now that we've got our linux mid server installed and configured it's time to start it up back on the instance we can see under mid server servers that our new mid server has registered but it still needs to be validated so we open up the record and click validate under related links we'll take the default options and click save the instance then tells the mid server to restart and is now ready to use let's try it on a data source we created earlier to get a list of employees from a mysql server behind our firewall here's where we put in the value for the mid server the rest of the fields are pretty much the same as they would be if we had a local mysql server in the interest of time for this demonstration we're going to limit it to about 10 records rather than the 300 000 that are available clicking through the transform map we see the field map coalesce value and one transform script that runs before each record is saved it takes the first name and last name fields from the source and saves it to the target name field when we go back to the data source and load some records we can see the instance working with the mid server to get the data the mid server then returns the payload and the records are saved in the import set table from here we'll run the transform and can see the test employees saved in the target table the import looks great just a little information about what's going on behind the scenes when a data source works with the mid server first the data source puts an output request into something called the eccu think of it as an inbox or outbox on your desk we just put a message into the outbox the eccq is the primary channel the mid server uses to exchange work with the instance the mid server makes an outbound request every few seconds to check for any work when it sees an output record in the ecc queue it picks it up and processes it after the mid server is done processing the request it makes another outbound request to send the data back to the instance this is an input record as if someone stopped by your desk and dropped a letter in your inbox the system knows that this input request is tied to the data source and processes it the same way as any other import running the transform and saving the records in the target table now you know a little bit more about what mid servers are how to get one running and what to do with it they can be very useful when you need to interact with data inside your organization's firewall