In a Webcast Wednesday, Cisco Systems Inc. (Nasdaq: CSCO) will try to convince the world (again) that it's already announced a game-changing data center fabric, countering the buzz that's lingered since Juniper Networks Inc. (NYSE: JNPR) unveiled its QFabric architecture in February.
Cisco has a few new products to show off as well, including new cards for the Catalyst 6500 -- still relevant in the Data Center 3.0 era! -- and the Nexus 3000 family of switches for high-frequency financial trading and other ultra-low latency applications.
It's also adding some cloud-related capabilities, such as the ability to do switching between virtual machines without getting a hypervisor involved.
Much of Cisco's emphasis, though, will be on its claim of sticking to standards with its fabric, as opposed to the proprietary connections in QFabric. Cisco also thinks its approach is more flexible -- supporting arbitrary network topologies -- and has "greater scale than our competitors are offering," says Craig Griffin, a Cisco product manager.
It's worth noting that when asked about competitors' fabrics, Cisco executives emphasize that 10,000 customers are using NX-OS, the operating system behind the Nexus switches. Of course, that's a byproduct of having gotten to market earlier -- Brocade One from Brocade Communications Systems Inc. (Nasdaq: BRCD) is new, and Juniper's QFabric isn't shipping yet.
Why this matters
It seems like Cisco had to say something in the wake of Juniper's QFabric launch. Its unified fabric was first announced in 2008, part of what Cisco termed Data Center 3.0. That message is temporarily being overshadowed by newer, glossier marketing.
Brocade, Cisco, Juniper and possibly others could spend years in a standoff over their competing data-center architectures. For now, the debate is at a big-picture level; it's going to be interesting to see how that changes as more details of other fabrics, particularly Juniper's, get revealed.
Overall I thought it was a good presentation, although I was a little surprised at how much thought and production went into it and the fact that they trotted out so many high level execs. Do you think they were caught off guard by how much attention Juniper got with QFabric?
Cisco does have a couple of interesting items that got announced.
LISP (Location/ID separation) is a standard going through the IETF. Cisco uses it to create separate address spaces for a physical device and its location - basically untethering the device, allowing an IP address to migrate around the cloud to where resources are available. Seems like a useful idea.
The idea of extending FEX (fabric extenders) down to the VM level seems useful too. Adds a deeper level of simple portability in the cloud.
Cisco always likes to talk about a standards-based approach, but often that's referring to Cisco-born technologies that Cisco is guiding through the standards process. I wonder how much of that is in its fabric technologies.
Cisco always likes to talk about a standards-based approach, but often that's referring to Cisco-born technologies that Cisco is guiding through the standards process. I wonder how much of that is in its fabric technologies.
"
The problem is that other competitors just follow Cisco's approach and has no innovation. They only started the product development after Cisco shipped their box. Looks at reality, which vendor has the product on market before spreading the buzz word? The real product matters, and the buzz words are just some add-on ads in telecom/datacom market.
scottd -- i agree with you that it was long. I didn't last through the whole thing.
I do think they were taken aback by the attention Juniper got, especially for a product that's not shipping yet (but at least it's close and does exist). Cisco also needed to refresh several aspects of its data center story -- they had no super-low-latency switch for the data center, for instance, and that's something that competitors would have been able to really hammer on this year.
quicktime -- I understand what you're saying, but neither Brocade nor Juniper had an Ethernet story before 2008, so by definition they'd be doing this after Cisco did.
And the flip-side analysis, according to some reports I've been reading since Wednesday, is that Cisco fell behind in the data center since 2008 -- the Nexus 7000 needed updating, and there was no ultra-low-latency switch in the story. In the latter case, in particular, when you ask which vendor has the produt on market first ... it wasn't Cisco.
Foundry having been around for years doesn't count for Brocade in terms of having an Ethernet heritage before 2008? Did Brocade fire everyone from Foundry?
Hmmm? MX an "Ethernet Product" ? Sure, in the same way as an M320, or a CRS, it has Ethernet ports. But don't tell me you would be proposing your customer to use an MX in the datacenter in the same way as a Nexus or a BigIron.
True, although the MX was originally focused purely on carriers and JNPR was somewhat late to the party on that one (although they made up ground quickly). Still, I think that experience and their routing heritage has played a key role in the company establishing itself as a top player in the data center. Bringing the same design principles they learned in the service provider space into their enterprise products allows them to tell a good story around TCO and keeps them out of the price-per-port discussion.
Unfortunately for Brocade I think many customers still view them as more of a Layer 2/commodity player (like Extreme).
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.
(10) | 
Digg
Del.icio.us
Reddit
Tweet This