The ever changing SAN
Our world is ever changing with new data, events or technologies challenging what we once thought to be gospel. Our 9th planet may now be a planetoid and just when I thought I was at peace with my knowledge of storage protocols and their respective places in the universe, along comes Fibre Channel over Ethernet.
Whether you are reading your favorite storage news rag or attending the latest SAN event, the industry is a-buzz with FCoE. Driven by the T11 group, this is billed as a solution to address convergence of the LAN/SAN in the data centers.
For organizations looking at more ways to be environmentally friendly, FCoE offers reductions in power consumption, cabling and cooling expenses while consolidating I/O traffic in server farms. FCoE should not require a dedicated Ethernet fabric and with pause based flow control, users should enjoy the same performance with FCoE that is currently provided by today’s FC SAN. Yes, FCoE is here to save the day as it claims to offer the same high performance of native FC, the opportunity to save cash and little disruption to current environments.
On the heels of 8Gb FC, FCoE is based upon 10Gb Ethernet and will require a new breed of NIC’s and switches to support it. Product availability and price per port will heavily influence the adoption of FCoE but this effort seems to be on the fast track for adoption.
FCoE has already stimulated an entire new line of product development from the top tier component manufacturers and the tool vendors that support them. The FCIA group even had a live FCoE demonstration at the spring SNW in Orlando.
We all want things easier and the notion of having a single network in the data center certainly has its appeal. However, few things are ever as simple as they appear and while FCoE may present benefits over the current architecture of data centers and storage, the users in the industry may take some convincing.
I was recently told “The future is Ethernet…Stupid!” I believe market forces will drive towards convergence and that FCoE will change the way we think about storage. Will it replace native FC? Not in the short term but the need for convergence is clear and the desire is strong, so get ready.
May 28th, 2008 at 2:30 pm
A number of industries such as finance,stock etc., depend on FC SAN’s for their daily business. I’m doubtful that FCoE will be readily accepted by these top tier organizations for many years.
R
May 28th, 2008 at 5:27 pm
Rob:
I completely agree that FCoE will need to be proven before it is placed in business critical environments. SAN managers in these industries rely on the stability, deterministic nature and familiarity of FC.
It may be that FCoE first finds a home in replicate sites or single application nodes. It may find its beginning in new installations.
The users of technology would like convergence - if it provides benefits. If it happens and how quickly it will depends on many factors. Assuming equal performance of FC, cost per port and integration will be primary items of concern.
Will FCoE happen? When might it take root? What conditions will help it become common place? Will it replace or run in parallel with FC as do other technologies?
Is the separation between LAN and SAN so big that it is destined for limited success?
Give us your thoughts.
May 29th, 2008 at 1:40 pm
I think there will always be a number of solutions in the storage market addressing the needs of various customers.
FCoE will likely be accepted by some while others will use FC, iSCSI or some other new storage protocol.
M
July 10th, 2008 at 2:55 am
yes, I agree
July 10th, 2008 at 2:55 am
oh yes
July 16th, 2008 at 6:40 am
yes i agree
July 16th, 2008 at 7:08 am
yes i agree