The big three suppliers in the SAN fabric box space are circling each other in a wary war dance but nothing much is actually happening. The 4Gbit/s FC transition is complete at Brocade, underway at McData and delayed at Cisco. McData has a 10Gig capability for ISLs. All three have WAFS initiatives. All three have intelligent switch initiatives. But what amongst these things will make a customer change its fabric box supplier?
Fabric box-using customers are cautious and conservative. Until the management and use of multi-vendor switches in one fabric is easy and transparent there is every encouragement to stay with a single switch vendor. FAIS and SMI-S will make the mix-and-match approach easier of course when they become real standards.
The fabric box market is approaching maturity I think. Most enterprises that could buy SANs have bought them. There is a second stuttering wave of storage consolidation going on which is bring storage from remote offices into the data centre and providing remote office access through WAFs. That will add more ports and switches into the fabric to cope with the extra traffic.
But it is a complicated move as it is bound up with data protection, WAN acceleration, data compression and the potential removal of servers completely from branch offices. These make the buying decision harder and there isn't a WAFS rush on at the moment.
There is another potential SAN traffic booster with iSCSI routers being used for FC SAN extension, products like the recent Emulex one.
The other potential market boost for SANS is in the low end. There we have low cost Fibre Channel SANs and iSCSI. If iSCSI wins out over Fibre Channel then that's storage networking traffic that isn't going through SAN fabric boxes, bandwidth denied to the fabric box vendors, such as Cisco.
Also the basic iSCSI appeal - it's the network everybody knows - is so powerful that unless iSCSI networks are bandwidth hogs and slow, slow, slow in everyday use it's hard to see FC SANS being that successful in the low-end.
Brocade has its WAFS and application management activities providing its strategy for boosting profit outside the maturing SAN fabric market. McData has survived the Cisco onslaught and built a SAN switch business as well as getting into iSCSI and WAFS and building its intelligent switch technology. It looks stable and capable of growth.
Cisco is the company in a hurry; after all, it signalled storage as a potential billion dollar business when it entered the fabric market, yet it's stopped being a technology leader. Where are the 4gig Cisco fabric boxes?
We might suppose that Cisco isn't going to get to a $1billion storage business with its current switches and directors. It needs to do something new. What might that be?
One obvious, to me that is, possibility, is to somehow embed iSCSI storage networking in its everyday, run-of-the-mill, Ethernet products. I'm thinking of some form of intelligence in them that manages iSCSI storage, rather like InVista in an intelligent fabric switch manages FC storage. That might provide the volume it needs.
Another possibility would be to buy Brocade or McData. But now I'm in speculation heaven and it's time to back out.
But with this last thought; energetic and aggressive vendors selling products into mature markets have to do something. Watch this space.