You know, I've been thinking about this for a while now, and I'm sitting in the IPAM webinar and the more I hear the more I think we need to see this feature in IPAM. The biggest problem I have with IPAM is how to structure or group my data in a way that is useful to me. By default in "Manage Subnets & IP Addresses", you get different groups like "Imported" and "Discovered" groups of IP addresses based on how solarwinds found out about them. Keeping them in these groups isn't too useful.
Under "Discovered" subnets I see groups of all the DHCP scopes named as our Microsoft DHCP admins named the scopes. This is definitely not very useful to me, and when you have more than X of these ( a limit you can define admittedly), you get a "X items reached" which if you click on it, you don't get it in the left-hand tree, but in the right hand pane. Navigating in the RH pane isn't that great IMHO...
Under "Imported" subnets I see groups of subnets that were imported (by IP) from my routers. I'm not positive, but I think that if the subnet is in the "Discovered" subnets groupings, it doesn't show up here.
Like most companies I've worked with the "master" list of IP addresses is still in a spreadsheet. Even though we've started trying to use IPAM, I don't see us moving away from that at least on a larger scale, because the spreadsheet allows us to structure things in the way we want to see it and because of that when either a range of IP's is allocated for a purpose or a specific subnet is assigned to a site, we can easily do this because the data in the spreadsheet is structured in such a way that makes it easy for us to get a higher level view of it. Not only that but if a large range has been allocated to a specific group or business unit, they can have their own spreadsheet from which they further allocate their IP's.
So, that brings us down to the feature I think we need to have in IPAM, and I think if done right, this will be a killer feature that everyone will love. And on the surface it seems to be a pretty simple feature. That feature is multiple views. And by that I mean multiple views that might show the same data in different ways.
Right now I'm all confused because I don't know what view will make the most sense. Do I want to see things based on their DHCP scopes? Or do I want it in a sequential format based on IP address? Right now you really only have 1 view, which you have choices in how to structure, but you really only get one choice in that structure. However, I really want to see my data in multiple ways, and I want an IPAM solution that adapts to those multiple ways! The problem now is if you put an IP or range in one group, you can't have it elsewhere. The layout you choose is your one view. I want multiple views that can all dynamically show my IPAM data.
So, one view might be an overview of all the higher end subnets that I get to play with. For me that would be like a spreadsheet with multiple tabs. On it I would see the major subnets I work with which includes not only our public IP's, but the private IP space also (10.X, 192.168.X, etc). Now, lets say I'm working with the 10.0/8 subnet, if I click on that I might want the next level of data that I see to be broken down into maybe the next octet of subnets (ie: /16's or 10.1, 10.2, 10.3 .. 10.255) or maybe even another way. Let's say I would then like the next level to be /9's (10.0, 10.64, 10.128, 10.192). The point being I'd like to be able to pick and manipulate my views. Within these views I would also like to see the DHCP data too when you get down to the actual IP levels.
Let's say we assigned a few ranges to one of our business units, lets say the 10.1/16 and 10.15/16 subnets. They further break those down into smaller subnets. I'd like to be able to assign them the ability to have views of both the 10.1/16 and 10.15/16 subnets, and potentially create their own views within that (or have me create them for them). But, in my views I still want to be able to see that I've assigned these to this business unit.
Maybe other views would be to see everything by business unit, or by the subnet size, what its used for (ie: server subnets, remote site subnets, Loopback addresses), or by which site its at, etc...
This is just an idea at this point, a rough framework. But I think if done right, it would be a killer feature. Rather than have one view, that might not please anyone, allow us to structure the data in a way that could replace our spreadsheets (which I believe most of us still use!).
Oh, another thing now that I think of it. Give us the option in these views to either obey subnet rules (ie: first IP is the subnet, last IP is the broadcast) or allow us to break them. Why? Let's say I allocated the 10.255.0.0/21 for loopback addresses. Right now I created a 10.0.0.0/8 supernet to hold the 10.255.0.0/21 supernet. Within that I break out the individual /24's just to group the loopbacks better. However, if I do that it shows IP's like 10.255.0.0 and 10.255.0.255, or 10.255.1.0 and 10.255.0.255 as unavailable because they're part of the subnet. However, I just want to use them for grouping, they aren't actual routable subnets on my network.
I'm sure this concept needs to be fleshed out a bit, but curious what others think?