Juniper announced their Junosphere emulator recently and have been smugly telling everyone how wonderful it is. Except it isn’t. Here is my take, FWIW.

SSL VPNs are not allowed

Junosphere uses an SSL VPN for access. Most companies don’t allow SSL VPN to exit their firewalls as standard security policy for data loss prevention. For EDU and ISP/SP this may not be true but for Enterprise (where Juniper is trying to get traction) it certainly is. Therefore, it’s not much use for corporate clients who have proper security policies in place.

Getting access

I’m not a reseller, not a customer, not a trainer – I’m a consultant who has limited options for accessing the product. You think I’m going to setup an account with a reseller, or that a reseller will even want my business ? There doesn’t seem to be any other way to buy it. What about companies who aren’t customers who want to prepare, dip a toe in ?

Think about Cisco customers who want to switch but are afraid of the operational impact of the CLI/Administration of JunOS. It’s a real problem. Personally, I’ve put millions of dollars of revenue into Cisco on the back of Dynamips proof of concepts, internal training programs and turned away JunOS because of lack of experience and exposure. You didn’t even get to the door.

The Value of Emulators in Change Management

Change validation – its hard to get changes approved these days. The ability to model a network in Cisco IOU or Dynamips (even unsupported or illegally) with 20 or 30 routers means people can validate complex MPLS/BGP/ISIS designs and changes that would otherwise never be allowed. And build customer confidence that it will work. Furthermore, their own people can access that content and develop their own skills. Brilliant.

This reinforces Juniper product in place. Makes it harder to switch away and creates sticky customers.

Nickel and Dimed

A customer who has spent millions on JunOS software, isn’t expecting to pay extra for something that should have been included since day one. A customer isn’t going to pay 5 bucks per day per router – it’s an extortionate rate. Plus it’s hard for companies to raise a PO for that in thousand dollars lots or variable amounts and worry about using it. I’ll Juniper employees couldn’t do it either.

Thats nickel and diming your customers. I would support a subscription program like MSDN – say hundred bucks and a license to keep the lawyers happy.

Sure, educational partners like what you’ve brought to market – they don’t have to build labs any more. No risk if it fails, no up front capital, no space, no business planning. Just hang up a shingle, market a course and deliver it. Costs are offset per student. Fine. Me ? I’m running a twenty router system validation that I get maybe two hours a day to develop while I’m juggling five major projects, five to ten staff and reports. Am I going to get USD$100 value per day ? No I am not.

The EtherealMind View

I’m not likely to start learning JunOS in Junosphere while it’s locked away in the cloud where I can’t use it when it’s needed, then I have to pay to use it, and it’s still not the way I want it or need it to be. Desktop or nothing. Everytime a consultant or reseller says “don’t use JunOS because its only used by service providers and ISPs”, you just lost a chance to open up an enterprise customer.

I’ve received a demonstration from Judy Beningson, Juniper’s vice president and general manager, Virtual Junos Business Unit. The product looks good, I like it. Clearly its early days and there is much more planned, but that’s OK because I wouldn’t expect anything less.

Junosphere is a chance to remove a key sales objection around product knowledge on JunOS. JunOS is the key marketing message for Juniper – one OS for all systems. Once customers are in, they can consider taking on other Juniper products. Goodness knows, many networkers are ready to consider alternatives to IOS. Why try and make beer money against a golden opportunity go and grab new customers in the Enterprise and bathe in champagne ?