we’d like to create group of endpoint using information extracted from a CMDB.
On the CMDB it is defined an information that allow to identify if a system is a test, certification or production system. This information it is not reachable from system properties and so the unique waty to have it is to get it from CMDB.
From CMDB we can, for example, make an export of a CSV file with “HOSTNAME, ENVIRONMENT”. Is it possible to import this file on IEM and to have a custom properties of the endpoint based on the value of the CVS?
I’ve ever worked for this before (just for design not for deployment implementation since my opportunity was slipped). In short I’m sure we can develop a way to do this.
In my case the suite data transaction directly between the customer’s CMDB and IEM was difficult, thus I planned the use of CSV file as you’re pointing. My solution design for them was that the IEM Client will receive the CSV file to the IEM Server in each specific period (e.g. daily) by the IEM policy action. The Client will evaluate in action whether or not information in the CSV does meet its inventory (e.g. the hostname can be found). After evaluation the file will be deleted at the end of action.
In addition to the design overview, my concern in that case was that no one could estimate so far what the size of CSV file and network traffic are like because the customer haven’t determined yet what kinds of information should be evaluated between CMDB and the IEM Client.
Hope that helps and any questions are welcomed (I hope I can answer it…).
As Toshiki’s solution suggests you need to get the information to your IEM endpoint so they can evaluate the property and return it to the IEM database.
It will also need to be on the client so it can be used for things like fixlet/task relevance, automatic computer groups, reports and so on.
You don’t put the information from your cmdb into the IEM database yourself.
There was another thread on this topic in the forums earlier this year but I can’t find it
I did something similar in our environment. I needed to make location specific settings for my point of sale systems, so I dump out those site specific values using SSIS and creating an XML file specific to that store’s location. Then I made a small FTP application that is in a policy and it pulls these settings down on that systems 1st check in. After that I used relevance and read the information I used pulled down from our FTP.
For example, to get the state that the computer is sitting in I use this example… while excluding servers I manage in the Datacenter that do not have this information…
“IF(Computer name contains “DCtr”) then “NA” ELSE ((node values of child nodes 0 of selects “StoreInfo/State” of xml document of file “C:\StoreInfo.xml”))”
I then can break down my properties and organize systems into automatic groups by state, size of store, number of POS systems, and fuel dispenser types. I suppose you could do the same thing with the CSV you are pushing from CMDB. XML works well though, from my experience, since there are inspectors to handle XML DOM.
