DATA CENTERS

  • 03/11/2010
    11:57 AM
  • Rating: 
    0 votes
    +
    Vote up!
    -
    Vote down!

Brocade Shifts Ethernet Strategy

I must admit I was surprised when Brocade, in their quarterly earnings call last month, announced that while their total revenue and earnings were up, Ethernet sales were down almost 25 percent. Analysts were busy downgrading the stock, declaring the Brocade/Foundry merger a failure and generally decrying the foolishness of anyone that dared challenge Cisco in the market. Methinks they doth protest too much.
I must admit I was surprised when Brocade, in their quarterly earnings call last month, announced that while their total revenue and earnings were up, Ethernet sales were down almost 25 percent. Analysts were busy downgrading the stock, declaring the Brocade/Foundry merger a failure and generally decrying the foolishness of anyone that dared challenge Cisco in the market. Methinks they doth protest too much.

What happened should have been predictable. Brocade, despite the brand recognition they have amongst storage geeks, is really an OEM company selling over 75 percent of their Fibre Channel gear with IBM, HP, HDS, EMC or Jeff's Storage Hut logos on it. They bought Foundry and tried to sell Ethernet switches the way they sell Fibre Channel switches, through OEM. They made OEM deals with IBM and Dell and were surprised the orders didn't flow in. Going OEM didn't generate big sales because, while storage guys buy storage networking, it turns out network people are an independent group. They won't buy IBM or Dell Ethernet switches just because the server guys tell them to.

The other problem with that approach is that Foundry was a high-touch provider. During the years I was building the Foundry based network at the college, we called our Foundry SEs for pre-sales, post-sales and emergency support.  They were an extension of our network team, and had expense accounts to cover the late night pizza. So when Brocade started treating their customers the way OEM suppliers treat end users, and trying to turn what used to be free support into professional services engagements, their customers were not the happiest of campers.

After conversations with some of my sources, I see that Brocade's management has let Foundry's people go back to what made them successful in the first place.  Making solid gear and supporting the heck out of it.  I hope they succeed.  If some version of Ethernet is going to take over the data center, there should be at least three viable Ethernet suppliers. Today, if you want to run FCoE you have four choices: Nexus 5000, Brocade 8000, Brocade FCoE blade in DCX, or if you're running IBM blades, the BNC switch with Qlogic FCoE module in the blade chassis. In my dreams, HP/3Com and Juniper come out with FCoE switches and FCoE actually means more choice for users not more vendor lock-in.

Random closing thought: Bob Metcalf, who invented Ethernet, once said "I don't know what kind of network we'll be using in 20 years but I know they'll call it Ethernet". When you compare DCB Ethernet with the old fashioned CSMA/CD Ethernet, you know he was right.


Log in or Register to post comments