I am sponsoring this case for Zhong Wang. It requests patch binding
and times out on 12/19/2008. It has been converted to a fast track
from a full case, so there are relevant materials in the case directory:
% ls -l inception.materials
-rw-r--r-- 1 markcarl sac 1151 Aug 14 09:30 fcadm.man
-rw-r--r-- 1 markcarl sac 19354 Aug 14 09:30 fcoe_initiator_20q.txt
-rw-r--r-- 1 markcarl sac 201425 Aug 14 09:30
fcoe_initiator_func_spec_v100.pdf
-rw-r--r-- 1 markcarl sac 7088 Aug 14 09:30
fcoe_initiator_onepager.txt
-rw-r--r-- 1 markcarl sac 208777 Aug 14 09:30
fcoe_target_func_spec_v103.pdf
-rw-r--r-- 1 markcarl sac 1374 Aug 14 09:30 fcoei.man
Template Version: @(#)onepager.txt 1.35 07/11/07 SMI
Copyright 2007 Sun Microsystems
1. Introduction
1.1. Project/Component Working Name:
FCoE (Fibre Channel over Ethernet) Initiator
1.2. Name of Document Author/Supplier:
Zhong Wang (zhong.wang at sun.com)
1.3. Date of This Document:
05/08/2008
1.3.1. Date this project was conceived:
Q2 CY07
1.4. Name of Major Document Customer(s)/Consumer(s):
1.4.1. The PAC or CPT you expect to review your project:
Solaris PAC
1.4.2. The ARC(s) you expect to review your project:
PSARC
1.4.3. The Director/VP who is "Sponsoring" this project:
Scott.Tracy at Sun.Com
1.4.4. The name of your business unit:
Software
1.5. Email Aliases:
1.5.1. Responsible Manager:
Roger.Dong at Sun.Com
1.5.2. Responsible Engineer:
fcoe-iteam at Sun.Com
1.5.3. Marketing Manager:
Margaret.Hamburger at Sun.Com
1.5.4. Interest List:
fcoe-interest at sun.com
2. Project Summary
2.1. Project Description:
This project provides FCoE (Fiber Channel over Ethernet) capability
to Leadville (the Solairs Fibre Channel initiator stack) by using
normal 10GbE NICs instead of special hardware adapters (CNA).
2.2. Risks and Assumptions:
The T11 FCoE standard is not finalized yet, it is still being revised.
But the most important parts (Working model, addressing, discovery,
frame format, etc) are steady now.
The schedule of this project could be affected by the availability of
hardwares (FCoE switches and CNAs) so that the interoperability
testing could be delayed.
3. Business Summary
3.1. Problem Area:
FCoE is a newly proposed T11 standard to transport Fibre Channel
traffic over 802.3 networks. It's advocated by all leading IT vendors.
Solaris has a good FC initiator stack but is lacking FCoE initiator
capability so far.
3.2. Market/Requester:
All customers that want their FC based datacenters to scale
cost-efficiently.
All customers who wants I/O consolidation in their datacenters.
3.3. Business Justification:
With FCoE, customers can achieve I/O consolidation by leveraging
their
existing FC investment and knowledge with the 10 Gigabit Ethernet
infrastructure. I/O consolidation results in less adapters, less
power
consumption, less cooling requirement and less cabling. Adding
support
of FCoE will enrich Solaris open storage stacks, make Solaris
competitive to other Operating Systems in terms of storage
connection
capability.
3.4. Competitive Analysis:
Cisco has open-sourced a software FCoE initiator stack on Linux.
Both QLogic and Emulex have announced CNAs (Converged Network Adapter)
which are hardware FCoE approaches.
Our approach is a software solution, but instead of inventing a new FC
stack, we will use the existing Leadville stack, as it's robust and
widely adopted. Existing Leadville users are able to use FCoE without
any major change in their experience.
Compared with hardware approaches (CNA), our software approach uses
much cheaper generic hardware (10GbE NIC) to get competitive
performance, especilly when running on modern multi-core systems.
3.5. Opportunity Window/Exposure:
The project will be integrated into OpenSolaris in Q1 CY09, after
integration of FCoE target.
3.6. How will you know when you are done?:
By installing the SUNWfcoei and SUNWfcoe package on Solaris system
with appropriate hardwares (supported 10GbE NIC and FCoE switch),
users can start configurating and accessing remote FC and FCoE
target
devices by using standard FC utilities (cfgadm, fcinfo).
4. Technical Description:
4.1. Details:
The FCoE initiator function will be implemented in two parts:
The upper part is a new kernel driver fcoei, and the lower part
is kernel driver fcoe (which will be delivered by FCoE Target
project first).
fcoei is a pseudo driver. For each network interface configured to
perform FCoE, a dedicate fcoei instance will be created. So there is
a 1:1 mapping between configured network interfaces and FCoE ports.
The FCoE ports work in the same way as normal FC ports for Leadville.
The interface between fcoei and Leadville is the standard FCA
interface, so fcoei is acting as a virtual FC HBA driver, it
transports FC frames to and from Leadville through the FCA inteface.
It also performs FC-2 functions like exchange management and sequence
management as a normal FC HBA does.
fcoe is the driver to receive/transmit FCoE frames via GLDv3 MAC
client interface. It also performs FCoE addressing, discovery and
frame encapsulation/decapsulation. fcoe driver is a single instance
pseudo driver, it glues multiple fcoei driver instances together and
provides common functions to fcoei driver.
Part of the source code in fcoei will be ported from www.open-fcoe.org
under BSD license.
4.2. Bug/RFE Number(s):
Not filed yet.
4.3. In Scope:
N/A
4.4. Out of Scope:
N/A
4.5. Interfaces:
Imported interfaces:
- Leadville FCA interface
- GLDv3 mac client interface
- fcoe client interface
Exported interfaces:
- ioctls of fcoei driver (project private)
4.6. Doc Impact:
New man pages: fcoei(7D)
Changed man pages: fcinfo(1M)
4.7. Admin/Config Impact:
New sub-commands will be added to fcinfo(1M).
4.8. HA Impact:
None
4.9. I18N/L10N Impact:
None
4.10. Packaging & Delivery:
A new package SUNWfcoei will be delivered.
4.11. Security Impact:
N/A
4.12. Dependencies:
This project has dependencies on:
- Leadville FCA interface
- GLDv3 mac client interface
- fcoe client interface
5. Reference Documents:
OpenSolaris web page for this project is at
http://opensolaris.org/os/project/fcoe/
6. Resources and Schedule:
6.1. Projected Availability:
Q1 CY09 (After FCoE Target project)
6.2. Cost of Effort:
Implementation & unit tests: 2 persons for 6 months.
Functional test: 1 person for 2 months.
Final documents: 1 person for 1 month.
6.3. Cost of Capital Resources:
Capital will be shared with FCoE Target project.
Cisco Nexus 5020 FCoE switch x 1
Finisar Xgig 10GbE analyzer blade x 1
SunFire T2000 x 1
SunFire X4200 x 2
SunFire X4500 x 1
Sun 10Gigabit Ethernet Adapter (Neptune) x 2
QLogic QLE8042 CNA x 1
Emulex LP21000 CNA x 1
6.4. Product Approval Committee requested information:
6.4.1. Consolidation or Component Name:
NWS Consolidation
6.4.3. Type of CPT Review and Approval expected:
Standard
6.4.4. Project Boundary Conditions:
N/A
6.4.5. Is this a necessary project for OEM agreements:
No
6.4.6. Notes:
None
6.4.7. Target RTI Date/Release:
Nevada build 112, Q1 CY09
6.4.8. Target Code Design Review Date:
10/01/2008
6.4.9. Update approval addition:
N/A
6.5. ARC review type:
Standard
6.6. ARC Exposure:
open
