I don't know about anybody else, but once again I find myself very frustrated. My concern is that you, that is SolarWinds, is continuing to concentrate its resources on bringing out new features at the expense of resolving existing issues.
Just as one example, I've been asking for years for business hour support in reporting. I've received nothing but platitudes that "it is being looked at" but version after version, nothing. If you guys can figure out how to make NetPath and now PerfStack work, surely you can invest a few dev hours and get business hours to work.
That is hardly the only example. There are many other issues that have apparently been put on ignore, and as a long time user I'm tired of waiting.
I cannot recommend that we continue to use SolarWinds until there is some movement on these older issues. I would hate to drop you, but I need this stuff fixed much more than I need a shiny new toy.
Thanks for the feedback, Ken. I will certainly take your points as they stand, as a plea for a higher ratio of improving on old content vs creating new.
Meanwhile, I can shed some light on our thinking in this area and how things may not always be what they seem. It really comes down to two things.
First, new and shiny features tend to be more interesting to talk about and hear about. Most people expect us to be progressing on old content all the time, and just want to hear about the new stuff. Clearly that is not true about the people in this thread. The result is that our messaging gives disproportionate attention to all the new stuff. An example of this is the fact that we had 84 bug fixes in this release, but that's not talked about almost at all. NPM's last release, which most of us have probably forgotten, was a Service Release and only included improvements to existing functionality. This doesn't mean we're working on improving old functionality as much as you would like, just that we're doing it much more than our messaging suggests.
Second, there seems to be an incorrect assumption about the difficulty of improving old functionality. Improving old content feels, as you mentioned, like it would take "a few dev hours". Unfortunately this is not true. The reality is that working on old functionality is substantially more time consuming than coming out with new features. The reason is that we have to be extremely careful to preserve existing functionality that we're upgrading, migrate data from the old to the new functionality, provide a path for users to move to the new feature, and handle all of the references in the code. We also have to be extra sure that it will not break. The day we release a new feature like PerfStack, no businesses depend on it for their day to day. Usage and dependence builds over time. Virtually everyone depends on unmanage, so improving it with the Mute option has several orders of magnitude more risk. Accordingly, we do dramatically more testing, automated, manual, and within customer environments. Now, all of this is worth it, which is why we invest so much in it anyway.
If you take those two into consideration and inspect the spread of content delivered in NPM 12.1, you may see it differently:
Mute - This is really an extension/improvement on the old functionality of Unmanage. This feature request took an amount of effort on par with PerfStack, but is not something we're talking about much. It's based on one of our most popular feature requests:
Meraki Wireless Monitoring - Although the backend is brand new, the intent was simply adding support for a new wireless controller in our old wireless monitoring functionality. It came from this feature request, among others:
84 bugs - Definitely improvements on old functionality, and took us more than little bit of time
Again, these points do not mean to suggest you're incorrect and I hear your request to focus more on existing functionality. I hope it does provide a view into our thinking.
At the end of the day, I agree that you guys deserve more improvements around core use cases. I will work to do better, and so will my teams.
First, new and shiny features tend to be more interesting to talk about and hear about. Most people expect us to be progressing on old content all the time, and just want to hear about the new stuff.
But you are not delivering on the 'progressing on old content' and that is what is bother your users -- UDT has not had a product update in 2 years, so it's hard to see what progressing on old content, or new features you have added in that time.
From my perspective, as I've gained more organizational responsibilities what I am hearing from my staff is how frustrating it is to use basic features, and so they are looking at products that make their life easier. As I have advanced I'm trying to spend less time doing basic alert configuration in solarwinds, and more time on the strategy for network infrastructure.
Richard Letts wrote:
But you are not delivering on the 'progressing on old content' and that is what is bother your users
I'd like to understand your view specifically better. Do you believe this is true of NPM, or UDT, or all Orion tools?
Migration of the server-based tools appears to have stopped. Some of them like the Trap Viewer, Syslog viewer, and Universal Device Poller are fairly significant because I have to give Server Admin roles to staff to run those tools. Note: I'm not even talking about changing the functionality of these parts significantly, I'm talking about giving a webUI with the existing functionality. UnDP and Atlas are slow to load and die unexpectedly with .NET errors. I avoid using them when I can and so we can't leverage as much functionality out of the product as we might.
It's somewhat ridiculous you can't set the Vendor and MachineType from an OID from the admin UI, I mean, looking at a thwack forum for you to add the name<>OID mapping to the base product. How does that make sense?
Aruba Wireless support is broken; it's been broken for years. I've given up trying to get it fixed.
VRF polling for UDT has been outstanding for 2 years. At one point you used to be able to set the L3 device and ROCID to get the ARP data for a subnet, but that functionality got stripped out (because it was hard for some people to use?).
Yep. One egregious situation I have developing today is having to give users Full Admin just to create pollers.
They are acting like I'm a jerk because I said no to full blown Admin access.
It's little stuff like this that makes my life painful.
How do I get around this?
Basic features such as discovery are so outdated and cumbersome to use. I can install any of the competing products and discover an entire class B network's devices in not much more time than it takes to go through that darn Wizard anytime you need to make a change
I can install any of the competing products and discover an entire class B network's devices in not much more time than it takes to go through that darn Wizard anytime you need to make a change. And then you have to break up multiple discoveries because SW can't handle a large range.
I can't just add an application monitor to a node when I am looking at it. I have to go back to the menu/applications assign it, etc.
Try to delete a ghosted drive or NIC. Can't do it by editing the node.
As was stated, UDT hasn't had an update in years. Same with NTD. I started with LANsurveyor, now on NTD which is less usable than LANsurveyor was. Try to export an updated map from UDT into Network Atlas. It duplicates it! Go through a wizard rediscover a range just to add a single node. And again, discovery times are excruciatingly slow. Pick any other discovery and mapping product out there and get a trial and try it out. It is magnitudes faster. Solarwinds just doesn't know how to do discoveries.
Reporting, while very flexible, has hardly any out of the box reports compared to others.
Solarwinds shines in flexibility, but apparently, no one at Solarwinds looks at their competitors, else they would see that they are falling behind in usability and out of the box experience.
dfairles I'm happy you're throwing down the gauntlet on the usability and out of box experience of the SW suite. You know why? Because we're definitely picking that gauntlet up. I think if you look at what's been happening with the releases starting with NPM 12 and continuing through to the latest release this year with PerfStack, you'll see that we're tackling those problems in each release. Consider this, less than one year ago we were still asking users to use a Win32 app to activate and manage licenses. Today, you can do all of these things through the web console. That's one of many projects that we've been working on behind the scenes to improve the usability of the product. These are not sexy features that we advertise, we're just making them happen, because we do care.
I would love to invite you to some of our UX research sessions. I'd also love to know if you're thinking of participating in any of the many betas that we are running at this time. As meech would say, THANK you for your feedback. I'm one of many Product Managers here that are listening.
Challenge accepted. I fully expect that you'll be happy with what we've got planned.
I would rather have automated, clean discovery of newly added or renamed disks rather than PerfStack any day.
I would rather have a non sluggish web console than Meraki monitoring. I don't use Meraki.
I would rather have a reliable CPU monitor that takes queue length into account than Arista monitoring. I don't have Arista devices.
I would rather have portable, copyable and customizable AppInsight templates.
I would rather have bulletproof server certificate inventory, reporting, management and monitoring than a silence alert feature.
The list goes on. Sorry to be grouchy. But I have a new client that is very frustrated as he discovers the major limitations and overall bugginess of his new system.
"I would rather have a non sluggish web console than Meraki monitoring. I don't use Meraki."
"I would rather have bulletproof server certificate inventory, reporting, management and monitoring than a silence alert feature."
These two right here are huge for me. I could not agree with you more on these points. Both have caused me plenty of stress and pain as of late.
We keep SCOM around for bulletproof certificate monitoring and reporting. I feel for solarwinds users that do not have it. Including my customers.
This is the kind of thing non-technical management & project manager people do not grasp. It keeps me up at night actually.
I was so relieved when we racked the ExtraHop appliance. It blows SolarWinds away when it comes to clicking a few times and getting all the certificate information you could ever need.
I feel like there are two camps -
Camp 1. SolarWinds is perfect, don't say it isn't, because it is. It runs flawlessly. It must be something you're doing wrong if it doesn't work.
Camp 2. SolarWinds is not running so smooth for us and has TONS of bugs, its frustrating, but we learn to live with it and try to ignore them as best we can.
So, which camp are you in?
We are losing our patience with around 5 glaring issues that are well documented and remain unfixed, but hey - look at all the great new features we have!
I love SolarWinds, but I can't make our directors, PMs, engineers love it when I can't run reports or do things that worked perfectly fine in 11.x
The tide is turning against SolarWinds at three of my largest clients.
I recognize what a massive piece of spaghetti code the product is, and I appreciate the complexity and work you guys put in. But the bugs are just killing us. Killing us with a vengeance.
I'm in Camp 3: Monitoring software is sort of like backup software. Which one do you hate the least?
At the end of the day, I just want something that is reliable and works. Bells and whistles are for attracting new customers. They know that is is difficult to switch once you have so much time and money invested.
They know that it is difficult to switch once you have so much time and money invested. I would compare it to cable companies that are only interested in attracting new customers rather than keeping their base.
I have come to the conclusion that a SolarWinds deployment with 5 in place upgrades in its footprint is trouble.
I have a brand new environment with all of the latest versions that runs great.
My larger clients with old builds that have been upgrade over several years have the big persistent problems.
It is too late though to convince the larger clients that SolarWinds is reliable. They simply don't want it anymore. New clients with new builds have zero issues. I wish old clients would let me rebuild and restore but they are demoing MaaS solutions and getting away from SW.
SolarWinds solutions are rooted in our deep connection to our user base in the THWACK® online community. More than 150,000 members are here to solve problems, share technology and best practices, and directly contribute to our product development process.