This Pretty Much Answers That Question

A couple of days ago I posed this question: is Sun preparing to take on Cisco? The question generated some interesting responses in the comments to the article.

Reader Bill had this to say:

How on earth would Cisco respond if Sun started introducing products with better performance, at a fraction of the price, built on high volume open source adoption?

As I responded, that’s the real $64,000 question, isn’t it? That’s the premise upon which this entire thing is built—that by using commodity hardware and open source components, Sun can produce high-quality, high-performing network equipment that they can sell for far less than Cisco.

Reader Ed, on the other hand, questioned the validity of this kind of move:

I would think that partnering with a Juniper or Foundry-type company and OEMing equipment from those companies would be a more prudent move than venturing on their own to create new network devices.

Normally, I would agree with Ed if we were talking about a company that was merely interested in entering a market in order to become a more complete supplier to their customers. That’s not Sun’s purpose. Sun’s purpose is, I think, to fundamentally change the nature of the networking hardware market. How successful they’ll be…well, that’s another question.

My original article also prompted a response elsewhere on the Internet. Christofer Hoff thought my use of the work “distracted” in describing Cisco and Project “California” wasn’t appropriate, and in one sense he’s correct—”California” is absolutely a natural evolution of Cisco’s products and technologies and it does make sense for them. As I pointed out to Hoff, though, being successful with this new solution (I can’t call it a server!) will take focus, and while Cisco is focused on “California” Sun has their opportunity.

And it looks like they are definitely going to take that opportunity:

As I’ve said before, general purpose microprocessors and operating systems are now fast enough to eliminate the need for special purpose devices. That means you can build a router out of a server – notice you cannot build a server out of a router, try as hard as you like. The same applies to storage devices.
 
To demonstrate this point, we now build our entire line of storage systems from general purpose server parts, including Solaris and ZFS, our open source file system. This allows us to innovate in software, where others have to build custom silicon or add cost. We are planning a similar line of networking platforms, based around the silicon and software you can already find in our portfolio.

The emphasis on that last sentence is mine, just to emphasize the clarity of where Sun is headed. Clearly, it is their intention to leverage OpenSolaris, Crossbow, ZFS, Solaris Zones, etc., to compete directly against Cisco. And Cisco appears to be their primary target, judging from this sentence:

That means you can build a router out of a server – notice you cannot build a server out of a router, try as hard as you like.

To me, that looks like a direct jab at “California”.

So, I guess the question of whether Sun is going to take on Cisco is settled. Hoff, get your popcorn!

Tags: , , , ,

  1. Daniel Reznicek’s avatar

    It would appear that “the cloud” is getting thicker. Redundancy and parallel processing by having the cloud model of many lower cost devices that re-route or takeover on failures is finally coming to fruition. The days of “routers route” and “servers serve” is blurring at break neck speed. Finally.

    I find it amazing that marketing convinces the masses that you have to have an appliance X that does Y when appliance X is clearly repurposed older technology with a nice looking cover. In many cases, I can buy at least 10 appliance X “innards” and build my own with Open Source or more affordable alternatives. I just have to do my diligence. Its not always a 1 to 1 feature set, but it will get me where I need to go 80% of the time. Sounds like Sun is painting that picture for us.

    If the architecture is built so that failures are mitigated by having services picked up by many different capable devices that don’t care if that service is a fileserver, a router, a whatever, you really have to ask yourself why the masses will keep buying the pretty box?

  2. Taylor Allis’s avatar

    Sun’s core assets have always been Solaris, processors, servers and the software that goes with it. Sun is not targeting Cisco directly per se – but what they call the entire “proprietary” IT world.

    Just as their world was disrupted by open software newcomer Linux and commodity processors; they now seek to disrupt markets with their version of open source software and commodity hardware.

    They aim to make routers, network devices, storage RAID arrays and controllers, etc. out of their OS and commodity hardware (processors, JBOD, SSD, etc.)

    One general purpose OS platform that can be re-swizzled a million different ways.

    Sun’s challenge will not be their technology or innovation – they have that in spades. It will be reining in their broad focus and general business acumen (partnership management, solid fiscal management, sales, marketing, distribution, etc.)

    Cisco may not have near the systems IP or portfolio as Sun, but what they do have is focus and business acumen…

  3. Dmitry’s avatar

    Remeber, back in the days, that’s exctly what people were doing – building routers and firewalls from commodity hardware, old servers, even workstations. Then Cisco came with an APPLIANCE and won the market. Same thing happened win NAS – NetApp won NAS marked introducing an APPLIANCE. So why enterprises will go 20 years back and start building routers and firewalls themselves from commodity hardware, instead of buying an APPLIANCE. The cost of the hardware is NOT the TCO as we kow, but just a portion of it (sometimes even smaller that support, maintenance, management, etc)?
    I’m with Christofer Hoff on the point that Cisco is after Unified Data Center, where integration is the key.
    Sun, on the other hand, is after students, hobbysts , enthusiasts innovators that have free access to open source OpenSolaris and can build something usefull out of it, so then Sun can sell them support and probably their hardware (as Sun can clain it is optimized for their software best of all). It reminds me RedHat model, but somewhat enhanced.

  4. slowe’s avatar

    Dmitry,

    Who said anything about people building these things themselves? The only thing we’ve said is that Sun is aiming to build networking equipment from commodity equipment using open source software–that doesn’t mean it won’t be a fully supported, pre-configured type of solution.

    Clearly, if Sun does this right, it will be pre-configured, easy to use, well documented, but very FLEXIBLE. If Sun does it wrong–and this is a very real possibility–then it will be poorly integrated, and your original question is very relevant: why? No one will bother, and Sun will fail miserably.

    It all falls on Sun’s ability to follow through and execute on the business plan.

  5. Marc Richards’s avatar

    Dimitry,

    They absolutely plan on building an appliance. That is exactly what they did with storage.

    Don’t count out the “enthusiasts innovators” either. I am sure there will be a couple startups out there who take the technology and do interesting things with it that Sun didn’t plan on doing…but that is good for Sun too. It creates disruption at the lower end of the market and it makes it easy to cherry pick good ideas down the road. What better way to enter a new market segment than to acquire some startup that uses all your technology?

  6. Nate’s avatar

    Yes commodity hardware is good nowadays, but is it really ASIC good? While many server folks would probably be ready to accept commodity server hardware as sufficient, I have to believe networking folks won’t cozy up to the idea as quickly. My guess would be that commodity hardware might get you through on some networks, but as networks evolve and demands increase will they be able to keep up? The current move is to 10gig. Juniper is already building their gear to support 100gig. Will Sun really be able to keep up at those speeds all in software?

    I have to agree with Dmitry a little bit that it does seem like a time warp idea. Granted the clarification that it is vendor built and supported commodity gear and open source software, we are still talking about the same idea. I guess Microsoft is putting thin client mentality back into 2008 R2 with Remote Desktop Services though, so maybe we are just in that ripe time to reinvent old ideas. Like you said they may fly or flop. It’s pretty hard to tell what is going to happen now with things like cloud computing ramping up and Microsoft redefining the corpnet and pushing things up the stack with IPv6, IPsec, and isolation.

Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>