If you remember, a while back Cisco bought the all-flash storage company WHIPTAIL. There has been a lot of press and a lot of discussion around Cisco’s motivation. But what has happened since then? Where are they going with it?
WHIPTAIL helped fill a gap in Cisco’s datacenter portfolio, storage. But it’s far more important than that and much more of a longterm strategic move. Cisco has, arguably, the best server platform in the business which some distinct advantages over other architectures. The goal is to leverage the software and other IP from the WHIPTAIL acquisition to very closely integrate UCS with flash storage and automate it with UCS Director. It’s not hard to imagine the possibilities when you start looking at what you can do with the combination of these products. They’ll be able to offer very fast access to storage across the fabric interconnects or a centralized flash caching solution. It’s intriguing, for sure. The sooner they start showing integration with UCS and UCS Director the better. That’s going to be their key differentiator in an increasingly crowded market.
But what about today? Where are we at? The WHIPTAIL products are now known as the Invicta line and so far the only real change is that the 2U appliance is now a C240M3 Cisco-made rack mount server. The larger Scaling System should start shipping soon.
Cisco is starting to talk to customers about the Invicta line and they are going through a bit of a learning curve. To be blunt, Cisco has never sold storage and they are learning it’s different than route/switch or collaboration. The storage market is very competitive and though from afar the products appear similar there are very important differences. The relationships you need are also different and in many cases storage is a “sacred cow” where reputation and past experience means a lot. Having said that, Cisco has great customer relationships and years of solid reputation behind them in other areas. That isn’t totally lost when you make a shift in to a new area. Cisco has done it before and I’m sure they’ll do it again.
They are also off to a bit of a slow start on the products. The current appliance offers very good performance for the money. It’s probably the least expensive all flash offering from a major manufacturer on the market right now, but it’s not perfect. The biggest compromise you make for the money is the lack of actual redundancy. Remember, the Invicta Appliance is a C240M3 running the RACERUNNER operating system. The C240M3 has a single motherboard and the appliance uses a single LSI controller.
Storage administrators expect redundancy in any enterprise storage array. Cisco should expect to run in to some pushback when offering a storage system with some single points of failure. But if they position the cheaper appliance in the right use cases and create reference architectures outlining the deployment of applications with application redundancy it might be very successful. Customers could purchase two or more appliances for a reasonable amount of money and get great performance if the can depend on application failover instead of hardware resiliency. Would be an interesting VDI use case with the right architecture. I’m anxious to see Cisco start releasing good reference architectures outlining how to use these storage arrays.
Soon the larger Scaling System will ship. The Scaling System is a multi-node “cluster”. It has two front-end “routers” that handle all I/O to end-hosts. Data is stored on backend “storage nodes” that can be scaled for capacity and/or performance. It’s all connected via an Infiniband fabric. It allows Cisco to offer a truly redundant and scalable system and compete much better with the other offerings on the market. As I said above, they aren’t shipping just yet but should be soon and we’ll have one in the Varrow lab right after launch so expect to hear a lot more on that.
Cisco will be successful with Invicta. Cisco is successful at whatever they really want to do. They understand the importance in this offering and how important of a pillar it is in their overall datacenter strategy. They may be off to a bit of a slow start while they figure out how to integrate and sell the current products while working to build the next generation, but this is all temporary. The real short term risk is how Cisco’s other storage-focused partners react to the Invicta push. I’ll save that discussion for a later post. No need to get myself in trouble today….