DX Infrastructure Manager

Expand all | Collapse all

Don’t ask what IT monitoring can do, ask how it can do it

  • 1.  Don’t ask what IT monitoring can do, ask how it can do it

    Posted 03-31-2015 11:48 AM

    It’s not what we do that separates CA UIM from the rest. It’s how we do it.  Check out the most recent video we created that helps explain the CA UIM difference: http://bit.ly/1GbYH6t



  • 2.  Re: Don’t ask what IT monitoring can do, ask how it can do it

    Posted 04-01-2015 02:02 AM

    I get it that this is marketing material and as such understand the nature of such things. However, having heard the sales pitches and seen the marketing materials a couple of times over the years and having actually worked full time with the product for some years, I'd like to leave some comments. I believe it should just highlight on what the community feels based on discussion and the ideas posted. Presentations like this do somewhat provoke that.

     

    Message Bus is the underlying technology and as such I think it's pretty great. However it's just the foundation, just one component onto which you need to actually build your monitoring, user interfaces etc. The "how" entails a lot more from gathering data to analyzing it to publishing it (be it an alarm or a graph). Moreover, "the how" is also in big part your development process and the general architecture into which you steer the product.

     

    I'll jump into the VMEESH thing launched here:

     

    Visibility - It's there

    Multitenance - The whole access control system needs to be revised in my opinion. I have a hard time calling it truly multitenant.

    Ease of use - I don't really think it's fair to compare it to installing apps. First, in your phone you got one interface to get the app and launch it. This is not the case with UIM - and you really need to know which interface to go to to do a specific thing. AC vs IM

    Extensibility - It's there, though I'd love to bicker about things such as documentation and training.

    Scalability - True the message bus works well there, but will all the "central processing" components that are being added?

    High Availability - I guess I'm not comfortable with the definition of High Availability here. That the message delivery system is failsafe doesn't make UIM Highly Available. The "UIM deployment reference architecture" in itself sounds interesting - where can I find this documented? Especially with recent few major versions, configuring the HA for ALL the core components has become a really confusing thing and the word I get mostly is that it can't even be done. Moreover UMP, which is maybe the primary interface these days, itself doesn't support HA on an agreeable level! True, you can add UMP instances to horizontally scale, but there is no (supported) method to connect it to rest of the UIM infrastucture in a highly available fashion.

     

    Also, I get very often asked "what can it do? can it do this?" and this far not once have I been asked "how does it do it?". So naturally, that is the question I most often need to ask myself.



  • 3.  Re: Don’t ask what IT monitoring can do, ask how it can do it

    Posted 04-01-2015 06:15 AM

    Thanks for the reply and honest feedback.  I want to take some time to review your points.  I'll also see about getting you the deployment architecture. 



  • 4.  Re: Don’t ask what IT monitoring can do, ask how it can do it

    Posted 04-02-2015 10:57 AM

    Update:  The UIM Deployment Reference Architecture documentation is being updated to include 8.1/8.2 updates.  Once completed I'll post it.