I would also like to add a touch of extra complexity/confusion as "Has Impact" is not only a type of relationship (available in the 'Type' drop down) but its also an attribute of any relationship (via the HasImpact drop down - which offers Yes/No options).
So therefore any of Joanne's types can have their HasImpact set to Yes.
Yes Geoff that is correct, I have use both the 'Has Impact' relationship & 'HasImpact' drop down, but when you look at CI Explorer I can not see any difference in the view to show it has impact.
Also with regards to relationships, there does not seem to be an out of the box relationship for '2 way relationships', ie 2 servers that pass data/information between themselves as not a Source-Destination Relationship. Also no how do you show a relationship that is a Push/Pull of data relationship.
The direction and style for relationship type can be adjusted by system admins from CI Explorer.
Launch CI Explorer for any CI->Click on Settings tool bar icon->Goto Line Settings tab -> Select a relationship type
Adjust properties - direction, color, thickness or style.
The Has Impact field doesn't has any affect in the CI explorer window. The CIs which are shown as (X) along with their icon are the ones having an Incident or Problem linked to it.
The Has Impact field plays a role in calculating Impact analysis from Change Management.
Goto Remedyforce console->Open a change with linked CIs->Agent tools Menu->Click on Impact analysis
Thanks for the update, I have looked at this and will have a play.
I am now a little concerned regarding the Has Impact Field, as we were hoping to move to Consoles for Change to use the Impact Analyst function, but looks like I will have to now manually update the 1000s of CI relationships with Has Impact Yes/No, which will take a lot of time and resources to do.
You may well find the following of interest - as it is very specific about the use of the HasImpact flag. It is a copy of the help screen that is associated with the new 'Impact Analysis' button that was recently added to Change Requests.
I have raised a question about the use of the term 'Parent CI' within in the context since it appears contradictory to the use of 'Source of' as the driver for impact.
The above also implies that all impact is expressed in terms of a source affecting a destination.
This causes issues for bidirectional impact (which is the source?)
It also causes issues for hierarchies that define physical containers as destinations - e.g servers within a rack. Since we wish to progress up the hierarchy (source to destination) to find the location of the rack but downward (contrary to the way impact is defined) to find what is impacted when a rack dies.
We have noticed that there is NO bi-directional option for "Has Impact". Does anyone else see this as shortcoming in the product?
We have applications with relationships to servers and each CI has an impact to the other. The server impacts the Application and the Application impacts the server. Bi-Directional.
I would like to see BMC fix this. Anyone else?
Tom Lange - Can you elaborate on the bi-directional requirement with your application-server example? I'm not understanding how the application can impact the server. Note: The Remedyforce CMDB is modeled after Atrium which also does not support bi-directional relationships.
John Fulton I wasn't able to reply directly to your comment for some reason, so I am replying here.
To answer your question about "how an application impacts a server?", let me first preface this by saying that I am approaching this with 15 years of systems support/engineering experience. I have also spent some time on the other side, in the application support/configuration arena. In that time, we were measured on server "uptime". We would always joke that we ensure 100% uptime right up until the customer loaded their application onto the servers. After that, all bets were off. We had some apps that ran for months without any issue, some with very high traffic. Others, well... they made the on-call rotation miserable. To give you just one example, we had an application that was so poorly written it was guaranteed to crash the server every two hours, the Web service would just stop responding. We sequestered it to it's own machine so that it didn't bring down other applications along with it.
This is why I say that this relationship should be bi-directional. The app and the server work together and each has an impact on the other. I've only given one example here but if you need more... I've got them.
Thanks for paying attention to this.
Tom Lange - What are you trying to illustrate with a bi-directional relationship in this scenario? I ask because it would be a manual process to create and maintain these relationships and there must be some conditions you must define to determine which applications are bi-directional. Maybe if you provide a few more examples, I may get a better sense of this need. If you prefer, I am open to a call to discuss this. Feel free to send me a few available dates and times and I will set up a call. Thanks
John Fulton, I'd love to talk about this over the phone. I'd like to talk about what we are trying to accomplish and discuss why I think this change would help us. I am available on Tuesday afternoon, as well as Wed afternoon. For this meeting I would also like to invite our Change Manager, Lynn to get her perspective as well.
I believe you have my email, correct?