Highlighted
Regular Contributor.
Regular Contributor.
502 views

uCMDB hierarchy

Jump to solution

Hello,

I want to create a pattern view report for our business services and I have a question about the hierarchy pane in the uCMDB.

My TQL Query starts with the business service and ends with virtual servers. The different types of business services also use different kinds of service trees and CIs, like SAP Systems, Oracle, Linux etc.

I want to keep the TQL as simple as possible, so only adding a Computer once would be preferable. I already did that in the TQL, but I cannot figure out this principle in the hierarchy pane. I can't use a type multiple times.

Let me give you an example for the TQL, the letters are types of CIs:

A --> B --> C --> F

A --> D --> E --> F

(A and F is the same type and one instance in the TQL)

F should be in the hierarchy unter C and under E and should only be one instance, so I don't have to add it multiple times and change the attributes every time.

Is it somehow possible to to that?

Thank you very much!

PA

0 Likes
1 Solution

Accepted Solutions
Highlighted
Acclaimed Contributor.
Acclaimed Contributor.

Re: uCMDB hierarchy

Jump to solution

Once you see how the templates work, you may find them very userful for your case, if you need just certain number of reports, which to be reused in the future. 

You have to set the filtering on attributes and define the filter as parameterized. Then when you save the view, it will become template, on which you can attach template based views based on a file import with values for the settings or manual input:

template1.PNG

 

This is the screen where you can create X ammount of views differentiated by the values of the filtering. The good part is that if you change the template, all the views can be updated automatically.

template2.PNG

UCMDB is made to be used from the interface and exporting the data into flat file is not the best option. 

 

Cheers,

Petko

Likes are appreciated!

View solution in original post

7 Replies
Highlighted
Acclaimed Contributor.
Acclaimed Contributor.

Re: uCMDB hierarchy

Jump to solution

Hi PA,

What are you looking for is called compount relationship. Instead of creating complex TQL, just fill the source and the destination and after selecting them, choose Compound Relationship with the right button of the mouse. There you can specify all the pairs:

A-->B

B-->C

C-->D

A-->D

D-->E

E-->F

Keep the maximum steps minimal not to load the system too much. 

comp.PNG

 

And if the answer is helpful, please leave a Kudo!

Regards,

Petko Popadiyski

Microfocus CMS UCMDB Consulting

Likes are appreciated!
0 Likes
Highlighted
Regular Contributor.
Regular Contributor.

Re: uCMDB hierarchy

Jump to solution

Hello popadiyski,

thank you for your reply!

The actual problem is not creating the TQL, but creating the hierarchy/report.

The method you described works fine for the TQL, but afaik, you cannot include (for your example) infrastructure service and running software into the hierarchy. (See attachment)

0 Likes
Highlighted
Acclaimed Contributor.
Acclaimed Contributor.

Re: uCMDB hierarchy

Jump to solution
Let me know if I got your request straight. You want a report which will have B under A, C under B and D under C, but in case there is a connection between A and D, you will want the D to be under A, so with E under D and finally reaching F. But what if A can reach F both through B and C and D and F? Do you want the F to be show under C and E at the same time? And A and F should be the same CI Type in the query, because you don't want to set the properties on two places? So the query would be A to B to C to A or A to D to E and again A? Can B be connected to E as well or only the two paths should be possible?

Sounds like Schroeder's cat experiment and Inception 2 movie in the same scenario, which I would love to watch! And it literally blows my mind how much time you can save from clicking once instead of twice of the properties of the tql classes!
Likes are appreciated!
0 Likes
Highlighted
Regular Contributor.
Regular Contributor.

Re: uCMDB hierarchy

Jump to solution

Let me know if I got your request straight. You want a report which will have B under A, C under B and D under C, but in case there is a connection between A and D, you will want the D to be under A, so with E under D and finally reaching F.

Yes exactly.

And A and F should be the same CI Type in the query, because you don't want to set the properties on two places?

Wrong. The 2 Fs are the same CI type.

And it literally blows my mind how much time you can save from clicking once instead of twice of the properties of the tql classes!

That is only an example with two instances where you would actually have to set the properties twice. In realily, you would have to set the properties like 20 times for every minor change. Let's say I have 5 different paths from the business service to a computer and going on from the computer, I have the Filesystem, the CPU and the IP adresses. It sounds stupid to me if you have to add those CIs 5 times in the hierarchy. Also for every setting/attribute change, you'd have to do it for every CI...

Here's the problem: in the TQL, you are able to have n upstream CIs and n downstream CIs but in the hierarchy, you can match one CI to only one Upstream CI.

 

0 Likes
Highlighted
Acclaimed Contributor.
Acclaimed Contributor.

Re: uCMDB hierarchy

Jump to solution

I am just trying to imagine how do you expect to represent 2 dimenstion graph into 1 dimension spreadsheet with rows, if you want to be able to put the hierarchy as many times one level under another. If you can come up with this idea, you have to suggest it to R&D!

For me it is far more easier to simplify the TQL into 1 possible path only. If you are not content with having Business Service -> Compound Relationship -> Computer and you want the middle levels, then add them with some generic CIs. For example Business Service -> Infrastructure Element (filtered by CI TYpe in the properties)-> Infrastructure Element (filtered by Ci Type in the properties) -> Computer.  Then you can put the hierarchy as you want, but of course you will be limited in the specific attributes  you can report on.

And if you have to change attributes in several places all the time, then maybe it is better to set them as parameterized values and initialize them when you open the view.  You can also create template TQLs as well. 

Likes are appreciated!
0 Likes
Highlighted
Regular Contributor.
Regular Contributor.

Re: uCMDB hierarchy

Jump to solution

As an explanation, here's how the graph looks like right now (see attachment).

From the application, it goes either to an oracle DB, SAP, SQL DB or App Server, and they all end up with a computer and their performance/characteristics (CPU/Backup/Fileshare Space). So when I want to add another specific attribute to the list for the report, I'd have to do it for every CI, although it's the same type.

About the generic type: I don't think that works since the tree depth is different from VWFS Application to the computer, for instance: the app server goes instantly from VWFS App to the Computer, oracle DB has more layers in between.

I might try it again with a compound relationship even with losing some CIs in the report.

And if you have to change attributes in several places all the time, then maybe it is better to set them as parameterized values and initialize them when you open the view.  You can also create template TQLs as well. 

I have to check that. Haven't tried it right now neither do I know how it works 🙂

Thank you!

0 Likes
Highlighted
Acclaimed Contributor.
Acclaimed Contributor.

Re: uCMDB hierarchy

Jump to solution

Once you see how the templates work, you may find them very userful for your case, if you need just certain number of reports, which to be reused in the future. 

You have to set the filtering on attributes and define the filter as parameterized. Then when you save the view, it will become template, on which you can attach template based views based on a file import with values for the settings or manual input:

template1.PNG

 

This is the screen where you can create X ammount of views differentiated by the values of the filtering. The good part is that if you change the template, all the views can be updated automatically.

template2.PNG

UCMDB is made to be used from the interface and exporting the data into flat file is not the best option. 

 

Cheers,

Petko

Likes are appreciated!

View solution in original post

The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.