Discovery and CI Relationship

Amit Sharma4
Mega Expert

Hi There,

 

When we perform discovery there are few relationships which are created automatically by discovery.

Trying to understand, how those relationships are created. Like what all probes / scripts are used for this and how it is decided, what type of a relationship it is...

Please guide.

 

Thanks

Amit

1 ACCEPTED SOLUTION

Hi Amit,

 

You will be able to find some of this information Under pattern designer - Discovery pattern

if you take F5 Big IP as example so under Pattern tab  - you will be able to find the relationship rules and how the relationship type has been defined

Thanks

View solution in original post

12 REPLIES 12

DaveHertel
Kilo Sage
Kilo Sage

Hi Amit -- Your question is very broad, so a general answer is appropriate... Discovery architecture uses a series of queries (called Probes and/or Patterns) to interrogate devices.  The responses to probes can do many things, but most commonly they will either A) Update CMDB/CI's and possibly add relationships between CI's.

Relationships are often built automatically from classifiers, such as this example of a MySQL Server process classifier.  In this case, the mysql process is detected, a MySQL Instance type CI is created AND a relationship of "Runs on" is created from the mysql instance to its host

find_real_file.png

 

The types of relationships, CI's created, processes, detected varies by technology being detected & interrogated during a Discovery job. ... Hardware/Software/Apps.  There is much, much more to Discovery, this is just a quick example.  You may wish to consider attending a Discovery Fundamentals class that covers the primary aspects of Disco, CMDB, creating relationships, etc.

Does this help?  Hope so

Amit Sharma4
Mega Expert

Hi Dave

 

Thanks for the response and this general idea is very well in the direction of my requirements,

I was not clear, which probes / classifiers are triggered when discovery defines relationships. 

So, for s/w relationships like MS SQL, Oracle MYSQL etc, are created by process classification, once the respective process is detected running on the CI. I got this.

I was trying to see in cmdb_rel_ci table  to understand this but did not have any clue. There are other parent child relationships and others like Allocated to, Connects to, used by etc.

How these are define? is probes which will be finding h/w related probes or there some other classifiers for this as well?

 

Please guide.

 

Thanks

Amit

Hi -- Is possible the other relationship/types you see in cmdb_rel_ci are coming from Discovery, but you can't assume that.  It could be other data integrations/sources providing those relationships too (depends on your environment).  Classifiers certainly create relationships for apps-to-hosts and other stuff.  Looking at the classifiers will tell you what type of relationships are automatically created.


To get a bit more clue, on cmdb_rel_ci table, in list view I'd add the createdBy and UpdatedBy fields.  Then when viewing relationship records in that table, are they created by the account(s) that your MID server uses to do its work?  i.e. is the MID server user account you use for Discovery, adding/updating these relationship records?  if so, thats a pretty strong clue 🙂 that Disco is creating those relations.   

As mentioned, other systems/sources/processes could also be creating these relations.  Using the created/updated fields can provide some insight.  Perhaps you have business rules or even admin/people creating these relations...

 

Does this help? I hope so 🙂

Hi Dave,

 

so i tried to see few classifiers to see how IP Connection :: IP Connection relationship is created. I tried to see what is happening at field level but unfortunately , i did not find any thing , which could tell how that relationship was created.

Could you please throw some light, if your time permits?

 

Thanks

Amit