HP IMC 5.1 SP1 Update

A few months ago I wrote a review of HP’s Intelligent Management Center (IMC). There were some things I liked, some things I didn’t, and a few bugs I came across. HP has just released 5.1 SP1, a service pack update to the existing version. As expected, it fixes a few issues, but it also introduces a few new features. I don’t want this to be a product advertisement, but I did moan about some bugs and offer some crappy workarounds. So, I feel it’s fair I mention things that have been fixed.

Firstly, a disclaimer: HP’s Community Marketing group gave me free tickets to attend HP Discover, held recently in Las Vegas. My company, which partners with HP and other vendors, paid for flights and accommodation. HP also gave me early access to the IMC 5.1 SP1 code to try out. No-one asked me to say nice things about their products, although I’m sure they wouldn’t mind. I didn’t get paid anything, but nor was I out of pocket.

Interesting (to me) Changes:

• Cisco SSH Backups now work! Well, mostly. There’s now an option to choose “SCP” for file transfer mode for devices. This will attempt to use SCP, and fall back to using SSH + expect scripts. Note that you may need to enable scp on your IOS (“ip scp server enable”) or ASA (“ssh scopy enable”) devices. This seems to work well for IOS and ASAs. Falling back to SSH still doesn’t work properly for ASA though. It grabs the config, but doesn’t clean up the "--- More ---" prompts properly. Find and edit CiscoASA_Cleanup_Parser_Script.pl to match the ASA version of the “More” prompt. I don’t know who’s worse – Cisco for not standardising something so simple, or HP for not picking up on it (again).
• Display Tiling – meaningless name, nice feature for creating customised “big screen” views. Plenty of products provide a reasonable interface for engineer use on the desktop, but they can be tricky or impossible to create useful views to go on a big screen. This gives you a Flash-based (Boo! Hiss!) interface for pulling together graphs, maps, etc, and sizing/placing them just the way you want on a canvas. You can even embed a movie, I guess for real smoke-and-mirrors type demos.
• KVM support for Virtual Network Management. This is a broader industry trend, where pretty much everyone is adding support for virtualisation platforms other than VMware to their toolsets. It’s been a big shift in the last year, and I thoroughly approve.
• RESTful API – there’s now an API that lets you programatically control pretty much anything in IMC. This means you can have some other system make calls to IMC to orchestrate changes across your network, regardless of underlying network vendor. This opens up a world of possibilities for people with specific itches, and the time/knowledge to scratch it. Check out this post by @netmanchris for more info.
• VANs, or Virtual Application Networks. This one didn’t make much sense to me when I read the marketecture, but after a few conversations, and a demonstration by @netmanles, I think I’m starting to get it. Here’s the gist as I understand it: right now, when application teams are deploying new VMs, they need to talk to the network engineers, find out what’s required, then the network engineers go and make (potentially) large numbers of changes to network devices (implementing ACLs, QoS policies, etc.) Those changes have to be made on a lot of devices, to be prepared for wherever the VM may be deployed to. As more VMs get deployed, more manual changes are required. It’s a pain for network engineers, as they’re getting involved frequently, and they may hold up application deployment. Well, what if, instead of setting up rules for every single VM, we could set up a network profile that VMware administrators could see in vCenter and attach to their VMs? Then when the VMware admin deploys a new VM, the relevant network devices automatically get configured with the right policy? They can then deploy more VMs, attach the policy, and the network gets re-configured, without needing to involve the network engineer again. It’s still limited in scope, but I like the concept, and I think it’s something to keep an eye on.
• Service Health Manager and Application Performance Manager modules let you check application health, as well as network stats. You can then pull together different stats to create all the pretty little dials and charts beloved of your senior management team. Be careful not to knock it though, those simple colourful charts can be what gets your next budget round approved!

I’ve now installed this in several environments, and so far all seems to be going well. If you’re an existing customer, you should probably upgrade. Note that if you’re doing a new install, you’ll need to get the previous version (E0202) and install this one on top of it. You don’t seem to be able to just do a clean install of this.