Suggest you read through the recent postings (newer than maybe September 2020) to get a feel for what people are saying and maybe look at the ideas section to see what people are asking for in the product.
From my perspective (full acknowledge that everyone is a little unique so my concerns may not apply) the benefit of UIM when we purchased it was it's flexibility. You can pretty much monitor anything - if there's a command or script you can run then you can probably be confident you can monitor it. And the nas has the ability to run scripts against the inbound messaging so you can programmatically fix/enhance/adjust things on the fly. This gives one a lot of control over the functionality of the product.
Now the bad - the behavior of the product management group is very rip-and-replace. Sometimes with no replace. As an example, take the sunset of Flash - rather than port the existing UMP produce to be fully Flash it was instead removed from the product. EMS was released as a supposed replacement for nas/alarm_enrichment though that hasn't really gotten any traction. The templating functionality for probe configurations has been through five different iterations now and still doesn't come close to the archive package functionality. The admin tool IM hasn't been touched software wise (except for adding IP6 support) in 8 years but it's still a better admin tool than the current iteration of the web based admin console.
I think much of the issue is that the product has maybe 20% the number of developers it needs which puts the product management people into the situation of not having the resources to fix the existing issues completely, not having the resources to keep updating the existing feature set to keep current with the products they monitor and also not having the resources to build something new.
I think that most would agree here that product releases are at best beta quality and that one can expect to see patches to a given feature over the following 18 months to fix issues. It is better than it was several years ago but not good. To put it in a little perspective, my central hub has 50 probes active on it - seven of those probes are hot fix releases. Two or three more are running older versions because of defects in the current version for which there is no fix. The hotfix thing is a double edged sword - it's nice to get things fixed outside the product release cycle but on the other hand it would be nice if the GA versions were correct to begin with.
I think (I keep repeating that just to denote the possible difference between my perception and someone else's) that the recent product directions are also driven by standard RFP/RFI types of questions where Broadcom wants to make sure they get a check mark in a box whether the feature is useful or not. It's kind of one of those things where resources get wasted chasing things that don't matter long term.
I could go on but hopefully this gives a feel for my experience. The bones of the product are good but there are days where I feel like Broadcom is trying to convert a reliable draft horse into a poodle and winding up with a pony with bows in its mane.
Original Message:
Sent: 08-30-2021 11:55 AM
From: Don Henderson
Subject: UIM Evaluation
We are conducting an evaluation of UIM for its functionality/capabilities of the application. We are wondering if we can get some good, bad and ugly from other end clients that have been using the system. We are running NetOps 20.X with Spectrum and Performance Center and looking for the additional benefits that UIM may bring as well as some of the challenges folks have encountered. We are covering Data Center devices , AWS, and Azure so looking at it from a broad perspective. We have deployed a number of probes (OOTB) and are also looking at anyone that may have developed probes and what the LOE may be.
Any info would be greatly appreciated
------------------------------
Don Henderson
------------------------------