The frame-filtering folly

Ashok Kumar asks whether McData's lawsuit against Brocade over a promising new technology is less about serving customers and more about what happens when individual companies put their own interests first.

3 min read
In a recently filed patent infringement lawsuit, McData alleged that Brocade's frame-filtering measurement product infringes one of its patents.

Lawsuits may have become part of the technology landscape, but in a larger sense, this one raises the not-so-subtle question of whether the Fibre Channel industry is interested in serving the customer or whether this is an example of individual companies more concerned about what best serves them.

McData apparently has neither interest in a cross-licensing agreement with Brocade nor any motivation to gain access to Brocade's intellectual property through legal wrestling. This is a lawsuit that McData plans to vigorously press to stop Brocade from shipping products and leave its rival without a second option.

The interesting thing about this action by McData is that it carries a high level of risk if it fails. Indeed, Brocade's OEM (original equipment manufacturer) customers are not likely to cut McData any favors if this legal spat only serves to screw up the industry more than it already is. After all, Brocade has seen to it that switch interoperability is still an issue, which at the end of the day is what this lawsuit is really about.

Here's the technical background to put this dispute in context.

Frame filtering for Fibre Channel storage area networks is a relatively recent development with promising potential. Without frame filtering, a Fibre Channel switch is purely a low-level networking device. The general idea is to give the switches a way to monitor the application-layer data in Fibre Channel frames and provide this information to SAN (storage area network) management software.

Brocade has already invested considerable engineering resources to develop its fabric OS. But without the ability to respond to application-layer data, the value of the fabric OS would be severely constrained.

There are no current management applications that take advantage of frame filtering. In that sense, the loss of frame filtering may not seem like such a big deal to Brocade. However, nearly everybody involved with SANs knows that management is the main obstacle to more widespread adoption of the technology.

One of the most difficult aspects of SAN management is in correlating storage application processes with network processes. Frame filtering provides an important link in achieving this type of cross-functional management. OEM customers of Brocade may feel that their own value-added business plans are at risk if they plan on exercising frame filtering. For instance, an adverse ruling against Brocade could hypothetically blunt IBM's push into SAN management software development.

Because Fibre Channel is a gigabit speed technology, frame filtering is most likely done in hardware using techniques such as content addressable memory within the switch ASIC chip. An injunction against Brocade in this suit could force the company to respin its ASIC, or find an outside supplier of Fibre Channel ports for ASIC chips. Rest assured, McData probably would not be a supplier.

Of course, any such third party replacement chip would also have to find another way to implement frame filtering that doesn't encroach on McData's patents.

How the interests of customers are at all served by any of this remains one of the bigger mysteries of this still-young year.