Decomission MidServers

Nic Omaha
Tera Guru

We have been on SN for almost 11 years now and have decommissioned quite a few products such as Discovery and Service Mapping. Now our infrastructure is wanting us to decommsiion a number of our mid servers as they think they are not being used. We do currently use orchestration and I know which mid-server is set to do this, and have a ldap connection to AD. I was curious if anyone had any suggestions on a approach to see what ones might be being used for something. I think we can decommission some of these old ones but trying to tread very carefully. 

 

Thanks!

3 REPLIES 3

Maik Skoddow
Tera Patron
Tera Patron

Hi

basically your question can only be answered by you anbd should be easily possible with the respective documentation you hopefully have.

But you are right: you really have to be careful with deleting of MID Servers as they might be used in many configurations. To be sure that they are not referenced any more I recommend using my feature 

Maik

Community Alums
Not applicable

Hi @Nic Omaha ,

Having two mid-servers on one host and the old version is not being used but consuming space. 

  1. Find where the old MID-Server is installed
    1. Go to Run and type services.msc
    2. Find the Service called "ServiceNow MID Server"
    3. Right-click on the service and click on the Properties tab.
    4. See the "Path to be executable"
      • That is the path where the Configuration files of MID-server are stored. See the following screenshot as an example:

         

      • SandeepDutta_1-1692284997582.png

         

  2. Go to the MID-Server installed folder, take a backup of the agent and wrapper logs if needed.
  3. Go to the bin folder and double-click uninstall.bat
  4. The service gets uninstalled and now you can delete the entire MID-Server folder to mitigate the space issue.

Here is a strategy I used in our scenario , prepared some Mids, Created cluster ( new and old) . Obviously the new MiDs on a different place. Each Mid we were going to delete had a stand by on cluster so worst case scenario if something goes unintentional. The other will take over. 

Not sure if this will help you.