Published: 12 Sep 2002
Cisco's long-awaited Andiamo group - which was rolled right back into Cisco the same day the products were announced - is now out of the closet. The products are awesome - at least on paper. So, now what?
First, the issues. Fundamentally, Cisco has built a platform of Fibre Channel (FC) Directors, ranging from small to large, that promise to incorporate Ethernet (iSCSI and FCIP) as well as a platform for storage services, such as virtualization. Cisco's motivation is either one or two of the following:
- Their intentions are pure. They recognize that the FC market isn't going away, and they view it as big enough to try to dominate, until the day comes when Ethernet owns everything.
- They have ulterior motives. This is just the first step and they intend on expanding the market to incorporate disk subsystems on the back end - albeit relatively dumb subsystems since they want the overriding intelligence to reside on the 9000 series of switches.
Cisco and the FC market
The issue here is in execution, not the products. To date, Cisco hasn't made any hay with its first storage acquisition, NuSpeed. The company has learned that just because of who it is doesn't necessarily guarantee success in fringe markets. Cisco will need to bring in a lot of actual storage talent in order to compete in this space.
If the high-end FC market is your target, this naturally makes Brocade, McData, and InRange your enemy. Brocade is dominant in the midmarket - thanks to storage OEMs EMC, Compaq, and IBM - but it's a fledgling in the high-end market. McData is the kingpin of the Director space, and InRange has made inroads there. McData was put on the map by EMC, and InRange by IBM.
But, in order to steal share and dominate, Cisco has two options as far as I can tell.
One, they can build out a significant direct sales force and channel operation designed to speak to the storage buyer - a place Cisco has never been. They leverage their senior management relationships and attempt to sway the storage buyer - who doesn't know Cisco - to move ahead with a centralization strategy where Cisco is the heart. This strategy works if the storage OEMs don't get in the way, or better yet, jump on board, which probably won't be happening anytime soon.
If the storage OEMs slow down Cisco by not playing with them, or by forcing Cisco through the same long qualification process they made the others go through, it could cost Cisco at least a year. If Cisco really wants to put muscle into this strategy, it has the staying power and cash to make it happen, but it won't be a slam-dunk by any means.
Secondly, it can play the relationship card. Cisco keeps its strategy of maintaining high-end relationships and puts the squeeze on senior management to let them know that most companies already spend a ton of money on Cisco infrastructure, and that the storage guys with their FC stuff is really just a small extension to what will become a real enterprise storage network the same way that long-ago workgroup LANs were tied together to become an enterprise LAN. It can then try to push the top down to consolidate vendors, assuming Cisco can do what the storage guy needs.
If Cisco can appease the vendor community by making them believe that Cisco has no interest in the disk side of things, the vendors could react in one of several ways.
They climb on board and ride Cisco's marketing wave. The disk vendors reap some profit from reselling switches, but it isn't core to their being. If the enterprise is OK with Cisco, why shouldn't the storage guys? They will fast track Cisco through their qualification processes as soon as Cisco can demonstrate that it has some traction.
Even if Cisco says they won't take the disk business, many of the companies will be paranoid - and rightly so - since Cisco is big enough to dramatically affect all of their businesses. Therefore, they never want Cisco in the middle of their worlds and are slow to qualify them. In the meantime, Brocade and McData benefit hugely as they once again become the underdogs that the storage vendors can use as leverage to solve the problems, on their own terms. Brocade isn't big enough to dictate terms to EMC, and that's the way they like it. If Cisco elects to take the second strategy and starts to either directly or indirectly put disk systems behind their switches, you can bet the storage players will rally behind their incumbent providers. That is enough to keep Cisco out of the market in force for at least a year.
Networking is easy, relatively. At least that's what the storage guys will tell the customer - storage is the final mile. The rules are different here, mistakes cost you millions. This argument will be enough to cause the consumer to pause and should be able to slow down the mighty Cisco.
What will the switch guys think? They will be hoping Cisco screws up and shows its hand wanting the whole enchilada. If that happens, they will find friends in the big storage players like never before.
The interesting thing to consider is: Will the industry react the way the ATM players did when Cisco bought StrataCom? Will Brocade join forces with Foundry? McData and Extreme? One has to assume that the other large Ethernet players will follow suit and enter the storage market, less concede to Cisco.
Cisco and the disk business
Cisco will have no choice but to build out a massive sales operation of intelligent storage-speaking people - most likely stolen from the other players - and immerse themselves in the nuances of the business. There's merit to the technical arguments between storage networking requirements and standard networking. Cisco will have to understand things pretty quickly that took some big companies a long time to figure out. This could come back to haunt Cisco if they choose this route - one that could lead to the entire storage market rallying against them, leading to a complete meltdown. This strategy will unify the rest of the market against the great aggressor, and force Cisco to spend much more time and money than it really has too.
The bottom line
While they may have grander aspirations, Cisco should keep them to themselves, aggressively attack the high-end FC market, make its bones there and then consider where to go next. They then should try to understand the business models of the storage vendors, and make them partners before they decide to fight the program.
This move has the ability to dramatically alter the landscape, and two years from now, we may be looking at a far different world in storage than we now have.
Now for something completely different ...
Storage management software is becoming real. We're tracking some hot ticket startups that look like they have cool stuff that people really may need, and the money to stick around for a while.
Examples include the new Storability software, Monosphere. It makes a brilliant heterogeneous clustered volume manager, and a bunch of awesome provisioning players including ProvisionSoft, InterSAN, and CreekPath. All these guys build stuff that actually does something, not just tells you how ugly you are.
Watch for a Q4 boom for storage vendors. IT folks will be cleaning out the budgets they have left, which is all of it I think, in order to keep next year's money flowing. Likewise, watch Q1 to be abysmal.