SAN Design Cisco
-
Upload
arun-sondhi -
Category
Documents
-
view
266 -
download
6
Transcript of SAN Design Cisco
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
BRKSAN-1121
Storage Area Networking Core Edge Design Best Practices
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Session BRKSAN-1121 Abstract
SAN Core-Edge Design Best Practices
This session gives non-storage-networking professionals the fundamentals to understand and implement storage area networks (SANs). This curriculum is intended to prepare attendees for involvement in SAN projects and I/O Consolidation of Ethernet & Fibre Channel networking. You will be exposed to the introduction of Storage Networking terminology & Designs. Specific topics covered include Fibre Channel (FC), FCoE, FC services, FC addressing, fabric routing, zoning, virtual SANs (VSANs). The session includes discussions on Designing Core-Edge Fibre Channel Networks and the best practice recommendations around them. This is an introductory session and attendees are encouraged to follow up with other SAN breakout sessions and labs to learn more about specific advanced topics.
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Chad Hintz
Technical Solutions Architect-Data Center/Virtualization
CCIE #15729
Routing & Switching, Security, Storage
Who Am I?
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
What Are Storage Area Networks
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
History of Storage Area Networks
Fibre Channel Basics SAN Design Requirements
Introduction to Typical SAN Designs Introduction to NPIV/NPV
Core-Edge Design Review
Recommended Core-Edge Designs for Scale and Availability Next Generation Core-Edge Designs
Session Agenda
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 6
History of Storage Area Networks
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Direct Attached Storage: DAS
DAS – Direct-Attach Storage
Dedicated High Speed Access Can’t share capacity
Difficult to share data
Difficult to manage
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Network Attached Storage: NAS
NAS
NAS – Network-Attached Storage
Centralized Storage Attached over LAN
File sharing
More efficient capacity usage
Performance limits usefulness of NAS – mainly used for file storage and low-end databases
LAN
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Storage Area Network: SAN
NAS
SAN – Storage Area Network Dedicated ‘Back End’ Network
Match DAS performance Capacity deployed and redeployed Centralized management Diskless servers – simplified management,
reduced power and cooling
SAN
LAN
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
SAN Protocols
Fibre Channel Fibre Channel is a gigabit-speed network technology primarily
used for storage networking FCoE
Fibre Channel over Ethernet (FCoE) is an encapsulation of FibreChannel frames over Ethernet networks.This allows Fibre Channel to use 10 Gigabit Ethernet networks while preserving the Fibre Channel protocol
iSCSI iSCSI is a TCP/IP-based protocol for establishing and
managing connections between IP-based storage devices, hosts and clients
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
History of Storage Area Networks
Fibre Channel Basics SAN Design Requirements
Introduction to Typical SAN Designs Introduction to NPIV/NPV
Core-Edge Design Review
Recommended Core-Edge Designs for Scale and Availability Next Generation Core-Edge Designs
Session Agenda
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 12
Fibre Channel Basics
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
SAN Components Servers with host bus adapters
Storage systems RAID JBOD Tape
Switches
SAN management software
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Types of Fibre Channel Switches
Redundancy/Services Cisco MDS 9000
MDS 9506, 9509, 9513 MDS 9200
MDS 91XX
Small/Medium Business Enterprise and Service Provider
FC Bladeswitch
Edge Modular Director
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
‘N’ port: Node ports used to connect devices to switched fabric or point to point configurations.
‘F’ port: Fabric ports residing on switches connecting ‘N’ port devices
‘L’ port: Loop ports are used in arbitrated loop configurations to build networks without FC switches. These ports often also have ‘N’ port capabilities and are called ‘NL’ ports.
‘E’ port: Expansion ports are essentially trunk ports used to connect two Fibre Channel switches
‘GL’ port: A generic port capable of operating as either an ‘E’ or ‘F’ port. Its also capable of acting in an ‘FL’ port capacity. Auto Discovery.
Fibre Channel Port Types
N N
N F
NL FL
L L
E E
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
G_Port
G_Port
E_Port
F_Port
F_Port
E_Port
N_Port
N_Port
Fibre Channel Port Types
Fibre Channel Switch
NPV Switch F_Port NP_Port Fabric
Switch
TE_Port Fabric Switch TE_Port
End Node
End Node
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Start from the Beginning… Start with the host and a target that need
to communicate Host has 2 HBA’s (one per fabric)
each with a WWN Target has multiple ports to connect
to fabric Connect to a FC Switch
Port Type Negotiation Speed Negotiation
FC Switch is part of the SAN “fabric” Most commonly, dual fabrics are
deployed for redundancy
FC
HBA
Core
Initiator
Target
FABRIC A
Edge
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
My Port Is Up…Can I Talk Now? FLOGIs/PLOGIs
N_Port
F_Port
FC
HBA
Core
Initiator
Target
E_Port
Step 1: Fabric Login (FLOGI) determines the presence or absence
of a Fabric exchanges Service Parameters with
the Fabric switch identifies the WWN
in the service parameters of the accept frame and assigns a Fibre Channel ID (FCID)
initializes the buffer-to-buffer credits Step 2: Port Login (PLOGI)
required between nodes that want to communicate
similar to FLOGI – transports a PLOGI frame to the designation node port
In p2p topology (no fabric present), initializes buffer-to-buffer credits
Edge
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Buffer to Buffer Credits Fibre Channel Flow Control
B2B Credits used to ensure that FC transport is lossless
# of credits negotiated between ports when link is brought up
# Credits decremented with each packet placed on the wire Independent of packet size If # credits == 0, no more packet
transmission # of credits incremented with each
“transfer ready” received B2B Credits need to be taken into
consideration as distance and/or bandwidth increases
*See reference slides for B2B-Distance calculations based on Line cards.
16
16
Packet
15
R_R
DY
Host
Fibre Channel Switch
16
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Over-Subscription FAN-OUT Ratio Over-subscription (or fan-out) ratio for sizing ports and links
Factors used Speed of Host HBA interfaces
Speed of Array interfaces
Type of server and application
Storage vendors provide guidance in the process
Ratios range between 4:1 - 20:1
FC
6 x 4G Array ports 3 x 8G ISL ports
Example: 10:1 O/S ratio
60 Servers with 4 Gb HBAs
240 G 24 G 24 G
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Fabric Name and Addressing: WWN
Every Fibre Channel port and node has a hard-coded address called World Wide Name (WWN)
During FLOGI the switch identifies the WWN in the service parameters of the accept frame and assigns a Fibre Channel ID (FCID)
Switch Name Server maps WWNs to FCID WWNN uniquely identify devices
WWPN uniquely identify each port in a device
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Switch Topology Model
Switch Domain Area Device
Fabric Channel ID (FCID) Fibre Channel Addressing Scheme
FC Fabric
Domain ID FCID assigned to every WWN corresponding to an N_Port
FCID made up of switch domain, area and device domain ID is native to a single FC
switch limitation of domain IDs in a single
fabric Forwarding decisions made on domain ID
found in first 8 bits of FCID
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
FSPF “routes” traffic based on destination domain ID found in the destination FCID
For FSPF a domain ID identifies a single switch The number of Domains IDs are limited to 239/75 (theoretical limited/tested and qualified) within the same fabric (VSAN)
FSPF performs hop-by-hop routing
FSPF uses total cost as the metric to determine most efficient path
FSPF supports equal cost load balancing across links
Fabric Shortest Path First (like OSPF) Fibre Channel Forwarding
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Repository of information regarding the components that make up the Fibre Channel network
Located at address FF FF FC (some readings call this the name server)
Components can register their characteristics with the directory server
An N_Port can query the directory server for specific information Query can be the address identifier, WWN and volume names for all SCSI targets
Directory Server/Name Server (Like DNS)
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Login Complete…Almost There Fabric Zoning
FC
FC Fabric
Initiator
Target
Zones are the basic form of data path security zone members can only “see” and
talk to other members of the zone devices can be members of more
than one zone Default zoning is “deny”
Zones belong to a zoneset Zoneset must be “active” to enforce
zoning Only one active zoneset per fabric
or per VSAN
zone name EXAMPLE_Zone * fcid 0x10.00.01 [pwwn 10:00:00:00:c9:76:fd:31] [tnitiator] * fcid 0x11.00.01 [pwwn 50:06:01:61:3c:e0:1a:f6] [target]
pwwn 10:00:00:00:c9:76:fd:31
pwwn 50:06:01:61:3c:e0:1a:f6
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Enhanced vs. Basic Zoning
Basic Zoning Enhanced Zoning Enhanced Advantages
Administrators can make simultaneous configuration changes
All configuration changes are made within a single session. Switch locks entire fabric to implement change
One configuration session for entire fabric to ensure consistency within fabric
If a zone is a member of multiple zonesets , an instance is created per zoneset.
References to the zone are used by the zonesets as required once you define the zone.
Reduced payload size as the zone is referenced. The size is more pronounced with bigger database
Default zone policy is defined per switch.
Enforces and exchanges default zone setting throughout the fabric
Fabric-wide policy enforcement reduces troubleshooting time.
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Enhanced vs. Basic Zoning
Basic Zoning Enhanced Zoning Enhanced Advantages
Managing switch provides combined status about activation. Will not identify a failure switch.
Retrieves the activation results and the nature of the problem from each remote switch.
Enhanced error reporting reduces troubleshooting process.
To distribute zoneset must re-activate the same zoneset.
Implements changes to the zoning database and distributes it without activation.
This avoids hardware changes for hard zoning in the switches.
During a merge MDS specific types can be misunderstood by non-cisco switches.
Provides a vendor ID along with a vendor-specific type value to uniquely identify a member type
Unique Vendor type
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Zoning is used to control access in a SAN Soft zoning
Enforced by name server query responses Name server sends membership list to N_Port N-port accesses members only
Hard zoning Enforced by hardware (forwarding ASIC) at wire speed pWWN, fWWN, FC_ID, FC_Alias
Zoning—Enforcement
Zone-1
Array
Zone-2
Array
Host FC
MDS MDS Host FC
Soft Zone Hard Zone
Zone-1
Array
Zone-2
Array
Host FC
MDS MDS Host FC
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Hard zone
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Soft zone
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Analogous to VLANs in Ethernet Virtual fabrics created from larger cost-
effective redundant physical fabric Reduces wasted ports of a SAN island
approach Fabric events are isolated per VSAN which
gives further isolation for High Availability Statistics can be gathered per VSAN Each VSAN provides Separate Fabric
Services FSPF, Zones/Zoneset, DNS, RSCN
Virtual SANs (VSANs) Virtual Fabric Separation
Physical SAN Islands Are Virtualized onto Common SAN Infrastructure
VSANs supported on MDS and Nexus 5000 Product lines A Virtual SAN (VSAN) Provides a
Method to Allocate Ports within a Physical Fabric and Create Virtual Fabrics
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
SAN Islands – Before VSANs
Production SAN Tape SAN Test SAN
DomainID=8 DomainID=7
SAN B DomainID=2
SAN D DomainID=4
SAN F Domain ID=6
SAN E DomainID=5
SAN C DomainID=3 SAN A
DomainID=1
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
SAN Islands – with Virtual SANs
Production SAN Tape SAN Test SAN
SAN B DomainID=2
SAN D DomainID=4
SAN F Domain ID=6
SAN E DomainID=5
SAN C DomainID=3
SAN A DomainID=1
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Hierarchical relationship— First assign physical ports to VSANs Then configure independent zones per VSAN
VSANs divide the physical infrastructure
Zones provide added security and allow sharing of device ports
VSANs provide traffic statistics VSANs only changed when ports
needed per virtual fabric Zones can change frequently
(e.g., backup) Ports are added/removed
non-disruptively to VSANs
VSANs and Zones—Complimentary
VSAN 3
Physical Topology VSAN 2
Disk1
Host2 Disk4
Host1
Disk2 Disk3
Disk6 Disk5
Host4
Host3
ZoneA
ZoneB
ZoneC
ZoneA
ZoneD
Relationship of VSANs to Zones
Virtual SANs and Fabric Zoning Are Very Complimentary
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Criteria for forming a PortChannel Same speed links Same modes (auto, E, etc.) and states Between same two switches Same VSAN membership
Treated as one logical ISL by upper layer protocols (FSPF)
Can use up to 16 links in a PortChannel (32 Gbps max)
Can be formed from any ports on any modules—HA enabled
Exchange-based in-order load balancing Mode one: based on src/dst FC_IDs Mode two: based on src/dst FC_ID/OX_ID
Much faster recovery than FSPF-based balancing Given logical interface name with aggregated
bandwidth and derived routing metric
Inter-Switch Link PortChanneling
E.g., 8-Gbps PortChannel (Four x 2 Gbps)
E.g., 4-Gbps PortChannel (Two x 2 Gbps)
A PortChannel Is a Logical Bundling of Identical Links
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
ISL = inter-switch link
PortChannel = E_Ports and ISLs Trunk = ISLs that support VSANs
Trunking = TE_Ports and EISLs
PortChannel vs. Trunking
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
History of Storage Area Networks
Fibre Channel Basics SAN Design Requirements
Introduction to Typical SAN Designs Introduction to NPIV/NPV
Core-Edge Design Review
Recommended Core-Edge Designs for Scale and Availability Next Generation Core-Edge Designs
Session Agenda
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 38
SAN Design Requirements
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
High Availability - Providing a Dual Fabric (current best practice)
Meeting oversubscription ratios established by disk vendors
Effective zoning Providing Business Function/Operating System Fabric
Segmentation and Security
Fabric scalability (FLOGI and domain-id scaling) Providing connectivity for virtualized servers
Providing connectivity for diverse server placement and form factors
SAN Design Key Requirements
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Fibre Channel SAN Transport and Services are on the same layer in the same devices
Well defined end device relationships (initiators and targets)
Does not tolerate packet drop – requires lossless transport
Only north-south traffic, east-west traffic mostly irrelevant
Network designs optimized for Scale and Availability
High availability of network services provided through dual fabric architecture
Edge/Core vs Edge/Core/Edge
Service deployment
Client/Server Relationships are
pre-defined
I(c)
I(c) T(s)
T2
I5
I4 I3 I2 I1
I0
T1 T0
Switch Switch
Switch
DNS FSPF
Zone RSCN DNS
FSPF Zone
RSCN
DNS
Zone
FSPF
RSCN
Fabric topology, services and traffic flows are structured
The Design Requirements Classical Fibre Channel
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
History of Storage Area Networks
Fibre Channel Basics SAN Design Requirements
Introduction to Typical SAN Designs Introduction to NPIV/NPV
Core-Edge Design Review
Recommended Core-Edge Designs for Scale and Availability Next Generation Core-Edge Designs
Session Agenda
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 42
Typical SAN Designs
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Collapsed Core Design
Servers connect to the Core switches Storage devices connect to one or
more core switches Core switches provide storage
services
Large amount of blades to support Initiator (Host) and Target (Storage) ports
Single Management per Fabric
Normal for Small SAN environments
HA achieved in two physically separate, but identical, redundant SAN fabrics
FC
Core Fabric A Fabric B
Core
SAN Design – Single Tier Topology
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
How Do We Avoid This?
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
“Core-Edge” Topology- Most Common
Servers connect to the edge switches Storage devices connect to one or
more core switches Core switches provide storage
services to one or more edge switches, thus servicing more servers in the fabric
ISLs have to be designed so that overall fan-in ratio of servers to storage and overall end-to-end oversubscription are maintained
HA achieved in two physically separate, but identical, redundant SAN fabrics
FC
Core
Edge
Fabric A Fabric B
Core
Edge
SAN Design – Two Tier Topology
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
“Edge-Core-Edge” Topology
Servers connect to the edge switches Storage devices connect to one or
more edge switches Core switches provide storage
services to one or more edge switches, thus servicing more servers and storage in the fabric
ISLs have to be designed so that overall fan-in ratio of servers to storage and overall end-to-end oversubscription are maintained
HA achieved in two physically separate, but identical, redundant SAN fabrics
FC
Core
Edge
Fabric A Fabric B
Core
Edge
SAN Design – Three Tier Topology
Edge Edge
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
History of Storage Area Networks
Fibre Channel Basics SAN Design Requirements
Introduction to Typical SAN Designs Introduction to NPIV/NPV
Core-Edge Design Review
Recommended Core-Edge Designs for Scale and Availability Next Generation Core-Edge Designs
Session Agenda
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 48
Introduction to NPIV/NPV
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
What Is NPIV? and Why? N-Port ID Virtualization (NPIV) provides a means to assign multiple
FCIDs to a single N_Port Limitation exists in FC where only a single FCID can be handed out per F-port. Therefore and F-Port can only accept a single FLOGI
Allows multiple applications to share the same Fiber Channel adapter port
Usage applies to applications such as Virtualization
Applica'on Server FC NPIV Core Switch
Web
File Services
Email I/O N_Port_ID 1
Web I/O N_Port_ID 2
File Services I/O N_Port_ID 3
F_Port
F_Port
N_Port
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
What Is NPV? and Why? N-Port Virtualizer (NPV) utilizes NPIV functionality to allow a “switch” to act
like a server performing multiple logins through a single physical link
Physical servers connected to the NPV switch login to the upstream NPIV core switch
No local switching is done on an FC switch in NPV mode
FC edge switch in NPV mode does not take up a domain ID
Helps to alleviate domain ID exhaustion in large fabrics
FC NPIV Core Switch
Eth1/1
Eth1/2
Eth1/3
Server1 N_Port_ID 1
Server2 N_Port_ID 2
Server3 N_Port_ID 3
F_Port
N-‐Port
F-‐Port
F-‐Port NP-‐Port
Applica'on Server NPV Switch
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
NPV Auto Load Balancing Automatic Balancing of Load on NP Links
Uniform balancing of server loads on NP links Server loads are not tied to any uplink
Benefit Optimal uplink bandwidth utilization
Blade 1
Blade 4
Blade Server Chassis
Blade 2
SAN
Balanced load on NP links
Blade 3
1 3
2 4
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
NPV Auto Load Balancing Automatic Distribution Loads onto NP Links
Re-distribution of the loads among available links When the failed link comes up
Potentially disruptive
Not a default behavior
One time or explicitly configurable using CLI command/DM
Original distribution not maintained
Benefits No “left-out” NP links
Optimal use of uplink bandwidth
Scheduled traffic disruption
Blade 1
Blade 4
Blade Server Chassis
Blade 2
SAN
Blade 3
Disrupted servers re-login on other uplink
1 2
3 4
npv auto-load-balance disruptive
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
F-Port Port Channel Enhance NPV uplink Resiliency
F-Port PortChannels Bundle multiple ports in to 1 logical link Similar to ISL portchannels in FC and EtherChannels in Ethernet
Benefits High-Availability- no disruption if cable, port, or line cards fail Optimal bandwidth utilization & higher aggregate bandwidth with load balancing
Storage
Bla
de S
yste
m
Blade 1
Blade 2
Blade N
F-Port Port Channel
F-Port N-Port
Core Director
SAN
Interface port-channel 1 no shut
Interface fc1/1 channel-group 1
Interface fc1/2 channel-group 1
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
F-Port Trunking Extend VSAN Benefits to Blades
F-Port Trunking Uplinks carry multiple VSANs
Benefits Extend VSAN benefits to Blade servers Separate management domains Traffic Isolation and ability to host differentiated services on blades
Storage
Bla
de S
yste
m Blade N
Core Director
VSAN 1
VSAN 2
VSAN 3
F-Port Trunking on
F-Port Channel
F-Port N-Port
SAN
NPV
Interface fc1/1 trunk mode on trunk allowed-vsan 1-3
Interface port-channel 1 trunk mode on trunk allowed-vsan 1-3
Blade 2
Blade 1
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Application Protection and Consolidation Using VSANs and F-Port Trunking
Bla
de S
yste
m Blade 1
Blade 2
Blade N
F-Port Portchannel with Trunking
F-Port N-Port
NPV
Security and Services Differentiation for blades
Better application protection and isolation
Ability to host different application on blades in different VSANs
ERP
WEB
ERP
Web
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
History of Storage Area Networks
Fibre Channel Basics SAN Design Requirements
Introduction to Typical SAN Designs Introduction to NPIV/NPV
Core-Edge Design Review
Recommended Core-Edge Designs for Scale and Availability Next Generation Core-Edge Designs
Session Agenda
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 57
Core-Edge Design Review
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
“Edge-Core” Topology- Most Common
Servers connect to the edge switches Storage devices connect to one or
more core switches Core switches provide storage
services to one or more edge switches, thus servicing more servers in the fabric
ISLs have to be designed so that overall fan-in ratio of servers to storage and overall end-to-end oversubscription are maintained
HA achieved in two physically separate, but identical, redundant SAN fabrics
FC
Core
Edge
Fabric A Fabric B
Core
Edge
SAN Design – Two Tier Topology
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Core-Edge Design Options
Consolidation options High density directors in core
Fabric switches at edge
Directors or blade switches on edge
Scalable growth up to core and ISL capacity
A
B A A
A
B
B
B
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
A B
Server Consolidation with Top-of-Rack Fabric Switches
2X ISL to Core at 10 G
32 Host Ports at 4 G
96 Storage Ports at 2 G
28 ISL to Edge at 10 G
14 Racks 32 Dual Attached Servers per Rack
AB
Top of Rack
Top of Rack Design
MDS 91XXs Ports Deployed:
Storage Ports (4 G Dedicated):
Host Ports (4 G Shared):
Disk Oversubscription (Ports):
Number of FC switches in the fabric
1200
192
896
9.3 : 1
30
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Server Consolidation with Blade servers
2X ISL to Core at 4G
16 Host Ports at 4G
120 Storage Ports at 2 G
60 ISL to Edge at 4 G
Five Racks 96 Dual Attached Blade Servers per Rack
A B
A B
Blade Servers
Blade Server Design Using 2 x 4 G ISL per Blade Switch;
Less cables/power
Ports Deployed:
Storage Ports (4 G Dedicated):
Host Ports (4 G Shared):
Disk Oversubscription (Ports):
Number of FC switches in the fabric
1608
240
480
8 : 1
62
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Explosion of Fabric/Blade Switch in the Edge
Scalability Each fabric/blade Switch uses a single Domain ID (Tested 75 per fabric)
Theoretical maximum number of Domain IDs is 239 per VSAN
Supported number of domains is quite smaller (depends on storage vendors)
Manageability More switches to manage
Shared management of blade switches between storage and server administrators
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
History of Storage Area Networks
Fibre Channel Basics SAN Design Requirements
Introduction to Typical SAN Designs Introduction to NPIV/NPV
Core-Edge Design Review
Recommended Core-Edge Designs for Scale and Availability Next Generation Core-Edge Designs
Session Agenda
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 64
Recommended Core-Edge Designs for Scale and Availability
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
High Availability - Providing a Dual Fabric (current best practice)
Fabric scalability (FLOGI and domain-id scaling)
Providing connectivity for diverse server placement and form factors
Meeting oversubscription ratios established by disk vendors
Effective zoning Providing Business Function/Operating System Fabric
Segmentation and Security Providing connectivity for virtualized servers
SAN Design Key Requirements
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
A B
N-Port Virtualizer (NPV) Reduces Number of FC Domain IDs
2 ISL to Core at 10 G
32 Host Ports at 4 G
96 Storage Ports at 2 G
28 ISL to Edge at 10 G
14 Racks 32 Dual Attached Servers per Rack
AB
Top of Rack Top of Rack Design
Fabric Switches in NPV mode
NPV wizard for the deployment
MDS 91xx in NPV mode
Ports Deployed:
Storage Ports (4 G Dedicated):
Host Ports (4 G Shared):
Disk Oversubscription (Ports):
Number of FC switches in the fabric
1200
192
896
9.3 : 1
2
NPIV Core
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
NPV Reduces Number of FC Blade Switches for Server Consolidation
2 ISL to Core at 4G
16 Host Ports at 4G
120 Storage Ports at 2 G
60 ISL to Edge at 4 G
Five Racks 96 Dual Attached Blade Servers per Rack
A B
A B
Blade Servers
Blade Server Design Using 2 x 4 G ISL per Blade Switch;
Less cables/power
Ports Deployed:
Storage Ports (4 G Dedicated):
Host Ports (4 G Shared):
Disk Oversubscription (Ports):
Number of fabric switches to manage
1608
240
480
8 : 1
2
NPIV Core
MDS 91xx in NPV mode
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
High Availability - Providing a Dual Fabric (current best practice)
Fabric scalability (FLOGI and domain-id scaling)
Providing connectivity for diverse server placement and form factors
Meeting oversubscription ratios established by disk vendors
Effective zoning Providing Business Function/Operating System Fabric
Segmentation and Security Providing connectivity for virtualized servers
SAN Design Key Requirements
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
SAN Design FAN-OUT Ratio Use port-channeling/trunking to enhance bandwidth available
between devices Factors used
Speed of Host HBA interfaces
Speed of Array interfaces
Type of server and application
Keep ISL Oversubscription ratio lower than Array oversubscription ratio
Ratios range between 4:1 - 20:1
FC
6 x 4G Array ports 3 x 8G ISL ports
Example: 10:1 O/S ratio
60 Servers with 4 Gb HBAs
240 G 24 G 24 G
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
F-Port Port Channel Enhance NPV Uplink Resiliency
F-Port PortChannels Bundle multiple ports in to 1 logical link Similar to ISL portchannels in FC and EtherChannels in Ethernet
Benefits High-Availability- no disruption if cable, port, or line cards fail Optimal bandwidth utilization & higher aggregate bandwidth with load balancing
Storage
Bla
de S
yste
m
Blade 1
Blade 2
Blade N
F-Port Port Channel
F-Port N-Port
Core Director
SAN
Interface port-channel 1 no shut
Interface fc1/1 channel-group 1
Interface fc1/2 channel-group 1
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
High Availability - Providing a Dual Fabric (current best practice)
Fabric scalability (FLOGI and domain-id scaling)
Providing connectivity for diverse server placement and form factors
Meeting oversubscription ratios established by disk vendors
Effective zoning Providing Business Function/Operating System Fabric
Segmentation and Security Providing connectivity for virtualized servers
SAN Design Key Requirements
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Zones/ZoneSet
Create Device-Alias for End Devices
Create a readable name for end devices tied to their PWWN
As device moves between VSANs their Device Alias stays the same
Create 2 Member Zones
Hardware zoning on MDS
Recommended to have more zones with 2 members in larger SANs
Single Management of Zones/Zoneset per Fabric
Use Distribute full Zoneset command per VSAN to keep from isolation in
Basic Zoning or use Enhanced Zoning.
If a device has different active members the ISL will become isolated
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
High Availability - Providing a Dual Fabric (current best practice)
Fabric scalability (FLOGI and domain-id scaling)
Providing connectivity for diverse server placement and form factors
Meeting oversubscription ratios established by disk vendors
Effective zoning Providing Business Function/Operating System Fabric
Segmentation and Security Providing connectivity for virtualized servers
SAN Design Key Requirements
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Virtual SANs
Consolidate SAN Islands into OS or Department VSANs
Reduction of SAN islands into a single Fabric while keeping Isolation
Example is to have Test, Development and Production in their own VSANs
Separate Tape or SAN extension VSANs
Security
Create Separate Administrative Roles per VSANs
Use TACACS+ for authorization and auditing of Switches
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
VM-Aware SANs
Support for Port WWNs for Virtual Machine Hosted on Blade Servers
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
It is important to follow the guidelines from the virtual machine vendors for assigning port WWNs to virtual machines. For example, VMware requires the use of Raw Device Mode (RDM) instead of Virtual Machine File System (VMFS) to get access to raw LUNs.
Using NPIV/Nested NPV with RDM we can give QOS, incident isolation (VSANS) and visibility into a Virtualized environment
Individual LUNS vs. shared with the use of RDM with NPIV/Nested NPV
VM-Aware SANS
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Summary of Recommendations
High Availability - Provide a Dual Fabric
Use of Port-Channels and F-Port Channels with NPIV to
provide the bandwidth to meet oversubscription ratios
Hardware zoning with single Zoneset management per
fabric is key (2 member zones is recommended-Hard
Zoning)
Use VSANs to provide OS or Business Function
Segmentation
Use NPIV/NPV to provide Domain ID scaling and ease of
management
Use of host level NPIV and Nested NPV to provide
visibility to Virtualized servers
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
A B
Recommended TOR Core-Edge Design
NPIV Core
14 Racks 32 Dual Attached Servers per Rack
AB
Top of Rack Top of Rack Design:
NPIV Core
Fabric Edge Switches in NPV mode
MDS Edge in NPV mode
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Recommended Blade Switch Core-Edge Design
NPIV Core
Five Racks NPV Edge 96 Dual Attached Blade Servers per Rack
A B
A BBlade Server Design Using Core-NPIV
Edge-NPV
Top of Rack Blade Switch Design:
NPIV Core
Edge Blade Switches in NPV mode
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
History of Storage Area Networks
Fibre Channel Basics SAN Design Requirements
Introduction to Typical SAN Designs Introduction to NPIV/NPV
Core-Edge Design Review
Recommended Core-Edge Designs for Scale and Availability Next Generation Core-Edge Designs
Session Agenda
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 81
Next Generation SAN Design Best Practices
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
10Gbps Ethernet
Lossless Ethernet Matches the lossless behavior guaranteed in FC by B2B credits
Ethernet jumbo frames Max FC frame payload = 2112 bytes
Fibre Channel over Ethernet What Enables It? Et
hern
et
Hea
der
FCoE
H
eade
r
FC
Hea
der
FC Payload
CR
C
EOF
FCS
Same as a physical FC frame
Control information: version, ordered sets (SOF, EOF)
Normal ethernet frame, ethertype = FCoE
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
FC Traffic FC HBA
Unified Fabric Why?
Fewer CNAs (Converged Network adapters) instead of NICs, HBAs and HCAs
Limited number of interfaces for Blade Servers
All traffic goes over 10GE
CNA
CNA
FC Traffic FC HBA
NIC LAN Traffic
NIC LAN Traffic
NIC Mgmt Traffic
NIC Backup Traffic
IPC Traffic HCA
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Today’s Unified I/O Architecture
Ethernet FC
LAN SAN B SAN A
Today I/O Consolidation with FCoE
SAN B LAN SAN A
FCoE
Nexus 5000/UCS
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Nexus 5K FCOE Switches Also Use NPV to Achieve Both Server and IO Consolidation
A B
Attached Servers per Rack
AB
Nexus 5K/2k in NPV mode
LAN Core
32 Host CNA Ports at 10G
Core connectivity using FC modules on N5K
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Cisco UCS Fabric Interconnects Also Use NPV
A B
UCS 6100 fabric Interconnect in NPV mode inside UCS 5100 blade chassis
LAN Core
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
UCS Core-Edge Design N-Port Virtualization Forwarding with MDS
F_Port Channeling and Trunking from MDS to UCS
FC Port Channel behaves as one logical uplink
FC Port Channel can carry all VSANs (Trunk)
UCS Fabric Interconnects remains in NPV end host mode
Server vHBA pinned to an FC Port Channel
Server vHBA has access to bandwidth on any link member of the FC Port Channel
Load balancing based on FC Exchange_ID
Per Flow
Loss of Port Channel member link has no effect on Server vHBA (hides the failure)
Affected flows to remaining member links
No FLOGI required
SAN B SAN A
Server 1 VSAN 1
vFC 1 vFC 1
N_Proxy
F_Proxy
N_Port
6100-A 6100-B
F_Port
vFC 2 vFC 2
Server 2 VSAN 2
vHBA 1
vHBA 0
vHBA 1
vHBA 0
VSAN 1,2
VSAN 1,2
F_ Port Channel & Trunk
NPIV NPIV
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Nexus 5X00 Core_Edge F_Port Trunking and Channeling Nexus 5000 access switches
operating in NPV mode
With NX-OS release 4.2(1) Nexus 5000 supports F-Port Trunking and Channeling on the links between an NPV device and upstream FC switch (NP port -> F port)
F_Port Trunking: Better multiplexing of traffic using shared links (multiple VSANs on a common link)
F_Port Channeling: Better resiliency between NPV edge and Director Core
No host re-login needed per link failure
No FSPF recalculation due to link failure
Simplifies FC topology (single uplink from NPV device to FC director)
Fabric ‘A’ Supporting
VSAN 20 & 40
F Port Trunking & Channeling
VLAN 10,50
VLAN 10,30
VSAN 30,50
Fabric ‘B’ Supporting
VSAN 30 & 50
VF
VN
TF
TNP
Server ‘1’ VSAN 20 & 30
Server ‘2’ VSAN 40 & 50
Nexus 5000 NPV
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Fibre Channel Aware Device FCoE NPV
FCF
What does an FCoE-NPV device do? ”FCoE NPV bridge" improves over a "FIP
snooping bridge" by intelligently proxying FIP functions between a CNA and an FCF
Active Fibre Channel forwarding and security element
FCoE-NPV load balance logins from the CNAs evenly across the available FCF uplink ports
FCoE NPV will take VSAN into account when mapping or ‘pinning’ logins from a CNA to an FCF uplink
Emulates existing Fibre Channel Topology (same mgmt, security, HA)
Avoids Flooded Discovery and Configuration (FIP & RIP)
Fibre Channel Configuration and Control Applied at the
Edge Port
Proxy FCoE VLAN Discovery
Proxy FCoE FCF Discovery
FCoE NPV
VF
VNP
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Servers, FCoE attached Storage
FCoE Multi-Tier Larger Fabric Multi-Hop Topologies
Multi-hop edge/core/edge topology Core SAN switches supporting FCoE
N7K with DCB/FCoE line cards MDS with FCoE line cards (Sup2A)
Edge FC switches supporting either N5K - E-NPV with FCoE uplinks to
the FCoE enabled core (VNP to VF)
N5K or N7K - FC Switch with FCoE ISL uplinks (VE to VE)
Scaling of the fabric (FLOGI, …) will most likely drive the selection of which mode to deploy
N7K or MDS FCoE enabled Fabric
Switches
FC Attached Storage
Servers
VE
Edge FCF Switch Mode
VE
Edge Switch in E-NPV
Mode
VF
VNP VE
VE
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• BRKCOM-2002-UCS Supported Storage Architectures and Best
Practices with Storage
• BRKDCT-1044-FCoE for the IP Network Engineer
• BRKSAN-2704- Storage Area Network Extension Design and
Operation
Reference Sessions
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Recommendations
• NX-OS and Cisco Nexus Switching (ISBN: 1587058928), by David Jansen, Ron Fuller, Kevin Corbin. Cisco Press 2010.
• Storage Networking Fundamentals(ISBN-10:1-58705-162-1; ISBN-13: 978-11-58705-162-3), by Marc Farley. Cisco Press. 2007.
• Storage Networking Protocol Fundamentals (ISBN: 1-58705-160-5), by James Long, Cisco Press. 2006.
Recommended Reading
Available Onsite at the Cisco Company Store
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• Receive 25 Cisco Preferred Access points for each session evaluation you complete.
• Give us your feedback and you could win fabulous prizes. Points are calculated on a daily basis. Winners will be notified by email after July 22nd.
• Complete your session evaluation online now (open a browser through our wireless network to access our portal) or visit one of the Internet stations throughout the Convention Center.
• Don’t forget to activate your Cisco Live and Networkers Virtual account for access to all session materials, communities, and on-demand and live activities throughout the year. Activate your account at any internet station or visit www.ciscolivevirtual.com.
Complete Your Online Session Evaluation
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 94
Visit the Cisco Store for Related Titles
http://theciscostores.com
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Thank you.
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 97
Reference Slides
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
NPV Traffic Engineering
• Allows user to select external interface per server interface
• Benefits Allows customized Bandwidth Management
Allows use of shortest path
Enables use of Persistent FCIDs
Blade 1
Blade N
Blade Server Chassis
Storage
Blade 2
….
Traffic-map
SAN
Customized Traffic Pattern
1 2
N
npv traffic-map server-interface fc1/2 external-interface fc1/1 npv traffic-map server-interface fc1/3 external-interface fc1/1 ……. npv traffic-map server-interface fc1/N external-interface fc1/5
Fc1/3 Fc1/4 Fc1/24
Fc1/1 Fc1/5
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Number of NPIV Logins: MDS 9200/9500
Type of Logins Verified Logins Logins per Port 126 (1) / 256 (2)
Logins per Line Card 400
Logins per Switch 2,000 Logins per physical fabric 10,000
These are the number of logins allowed on all Gen1, Gen2 and Gen3 line cards. The limits applied to on a per switch will also apply to all MDS 9200 and MDS 9500. MDS 9124/9134 and Blade switches will have different limits and will be shown later.
(1) SAN-OS 3.x, NX-OS 4.1(1) (2) NX-OS 4.1(2)
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Number of NPIV Logins: MDS 9124/9134 and Blade Switches
Switching Mode NPV Mode
Logins per Port 42 (1) / 89 (2) 114
Logins per Port-Group 168 114
Logins per MDS 9124 1008 684
Logins per MDS 9134 1680 1140
Logins per MDS 9124e 1008 684
Logins per IBM Blade Switch 840 570
The stated numbers are verified / supported number of logins.
(1) Using 2 member zoning (2) Using default zone-permit instead of zoning
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
MDS 9000 Family Buffer to Buffer Credits Shared Dedicated
FX-Port (Fixed) FX-Port
(Default)
E-Port (Default)
FX-Port / E-Port
(Min-Max)
Extended Credits
(Min-Max)
Speed (Gbps)
Max Distance (*)
(km) Disable Ports for Max Credits?
18/4 Port 1/2/4-Gbps 16 16 250 2-250 256-4,095
1 8,190
No 2 4,095
9222i Fabric Switch 4 2,047
4 Port 10-Gbps N/A 16 750 2-750 751-4,095 10 683 No
4/44 Port 1/2/4/8-Gbps
32 32
125 2-250 251-4095 1 8190
No 2 4095
48 Port 1/2/4/8-Gbps 250 2-500 501-4095
4 2047
24 Port 1/2/4/8-Gbps 500 8 1023
9148 8G Fabric Switch N/A 32 32 1-125 N/A
1 250
No 2 125
4 62
8 31
9124/9134 Fabric Switch
N/A 16 16 1-61 N/A
1 122
No 2 61
4 30 (*) Assuming max frame size
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 102
Introduction to FCOE
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• 10Gbps Ethernet
• Lossless Ethernet Matches the lossless behavior guaranteed in FC by B2B credits
• Ethernet jumbo frames Max FC frame payload = 2112 bytes
Fibre Channel over Ethernet What Enables It? Et
hern
et
Hea
der
FCoE
H
eade
r
FC
Hea
der
FC Payload
CR
C
EOF
FCS
Same as a physical FC frame
Control information: version, ordered sets (SOF, EOF)
Normal ethernet frame, ethertype = FCoE
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Unified Fabric IEEE DCB
Standard / Feature Status of the Standard IEEE 802.1Qbb Priority-based Flow Control (PFC)
In Sponsor Ballot
IEEE 802.3bd Frame Format for PFC
In Sponsor Ballot
IEEE 802.1Qaz Enhanced Transmission Selection (ETS) and Data Center Bridging eXchange (DCBX)
Just completed WG recirculation ballot. New recirculation expected next week in order to go to Sponsor Ballot after the May interim
IEEE 802.1Qau Congestion Notification Done! IEEE 802.1Qbh Port Extender In its first task group ballot
Developed by IEEE 802.1 Data Center Bridging Task Group (DCB)
All technically stable
Final standards expected by mid 2010
CEE (Converged Enhanced Ethernet) is an informal group of companies that submitted initial inputs to the DCB WGs.
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
FC Traffic FC HBA
Unified Fabric Why?
• Fewer CNAs (Converged Network adapters) instead of NICs, HBAs and HCAs
• Limited number of interfaces for Blade Servers
All traffic goes over 10GE
CNA
CNA
FC Traffic FC HBA
NIC LAN Traffic
NIC LAN Traffic
NIC Mgmt Traffic
NIC Backup Traffic
IPC Traffic HCA
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
What’s the difference between DCE, CEE and DCB ?
All three acronyms describe the same thing, meaning the architectural collection of Ethernet extensions (based on open standards)
Cisco has co-authored many of the standards associated and is focused on providing a standards-based solution for a Unified Fabric in the data center
The IEEE has decided to use the term “DCB” (Data Center Bridging) to describe these extensions to the industry. http://www.ieee802.org/1/pages/dcbridges.html
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Priority Flow Control Fibre Channel over Ethernet Flow Control
Packet
R_R
DY
Fibre Channel Transmit Queues Ethernet Link Receive Buffers
Eight Virtual Lanes
One One
Two Two
Three Three
Four Four
Five Five
Seven Seven
Eight Eight
Six Six
STOP PAUSE
B2B Credits
Enables lossless Ethernet using PAUSE based on a COS as defined in 802.1p
When link is congested, CoS assigned to FCoE will be PAUSEd so traffic will not be dropped
Other traffic assigned to other CoS will continue to transmit and rely on upper layer protocols for retransmission
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
DCB “Virtual Links” An Example
VL1 VL2 VL3
LAN/IP Gateway
VL1 – LAN Service – LAN/IP VL2 - No Drop Service - Storage
Ability to support QoS queues within the “lanes”
Campus Core/ Internet
Storage Area Network
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Fiber Channel over Ethernet Protocol Host Side – FIP and DCBX Configuration
1st portion of the MAC is the FC-MAP of the Nexus 5000
FC-MAP (0E-FC-xx)
FC-ID 7.8.9
FC-MAC Address
FC-MAP (0E-FC-xx)
FC-ID 10.00.01
2nd portion of the MAC is the FC-ID
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
FCF (FCoE Forwarder): A Fibre Channel switching element that is able to forward FCoE frames (Nexus 5000, Nexus 7000, MDS 9000)
FPMA : ”Fabric Provided MAC Address” -- A unique MAC address that is assigned by an FCF to a single Enode
Enode : “End Node” -- a Fiber Channel end node that is able to transmit FCoE frames using one or more ENode MACs.
FCoE Pass-Through : any DCB device capable of passing FCoE frames to an FCF FIP Snooping Bridge FCoE N-Port Virtualizer
Single hop FCoE : running FCoE between the host and the first hop access level switch
Multi-hop FCoE : the extension of FCoE beyond a single hop into the Aggregation and Core layers of the Data Centre Network
FCoE Building Blocks The Acronyms Defined
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• FCF (Fibre Channel Forwarder) is the Fibre Channel switching element inside an FCoE switch
Fibre Channel logins (FLOGIs) happens at the FCF
Consumes a Domain ID
• FCoE encap/decap happens within the FCF Forwarding based on FC information
Eth port
Eth port
Eth port
Eth port
Eth port
Eth port
Eth port
Eth port
Ethernet Bridge
FC port
FC port
FC port
FC port
FCF
FCoE Switch FC Domain ID : 15
FCoE Building Blocks Fibre Channel Forwarder
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Fibre Channel Drivers
Ethernet Drivers
Operating System
PCIe
Ethernet
Fibre Channel
10GbE
10GbE
Link
Ethernet Driver bound to Ethernet NIC PCI address
FC Driver bound to FC
HBA PCI address
Replaces multiple adapters per server, consolidating both Ethernet and FC on a single interface
Appears to the operation system as individual interfaces (NICs and HBAs)
First Generation CNAs from support PFC and CIN-DCBX
Second Generation CNAs support PFC, CEE-DCBX as well as FIP Single chip implementation Half Height/Length Half power consumption
FCoE Building Blocks Converged Network Adapter
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Translate to FCoE… Same host to target communication
Host has 2 CNA’s (one per fabric) Target has multiple ports to
connect to fabric Connect to a capable switch
Port Type Negotiation (FC port type will be handled by FIP)
Speed Negotiation DCBX Negotiation
Access switch is a Fibre Channel Forwarder (FCF)
Dual fabrics are still deployed for redundancy
FC
CNA
FC Fabric
ENode
Target
Ethernet Fabric
DCB capable Switch acting as an FCF
Unified Wire
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
VE_Port
VF_Port
VF_Port
VE_Port
VN_Port
VN_Port
Fibre Channel over Ethernet Port Types
Fibre Channel over Ethernet Switch
FCoE_NPV
Switch VF_Port VNP_Port FCF
Switch
End Node
End Node
FCoE Switch : FCF
**Available NOW
**E Rocks Timeframe **EagleHawk + Timeframe
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
My Port Is Up…Can I Talk Now? FIP and FCoE Login Process
VN_Port
VF_Port
FIP Discovery
E_ports or VE_Port
Step 1: FIP Discovery Process enables FCoE adapters to discover
which VLAN to transmit & receive FCoE frames
enables FCoE adapters and FCoE switches to discovers other FCoE capable devices
verifies Lossless Ethernet is capable of FCoE transmit
Step 2: FIP Login Process Simular to existing Fibre Channel Login
process - sends FLOGI to upstream FCF adds the negotiation of the MAC address
to use Fabric Provided MAC Address (FPMA)
FCF assigns the host a Enode MAC address to be used for FCoE forwarding
FC
CNA
FC or FCoE Fabric
Target
ENode **Multi-hope FCoE with VE_Ports not supported until Eaglehawk Release
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 116
Enode MAC Address Fibre Channel over Ethernet Addressing Scheme
Enode MAC assigned for each FCID Enode MAC composed of a FC-MAP and FCID
FC-MAP is the upper 24 bits of the Enode’s MAC
FCID is the lower 24 bits of the Enode’s MAC
FCoE forwarding decisions still made based on FSPF and the FCID within the Enode MAC
FC Fabric
Domain ID
FC-MAP (0E-FC-xx)
FC-ID 7.8.9
FC-MAC Address
FC-MAP (0E-FC-xx)
FC-ID 10.00.01
Fibre Channel FCID Addressing
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Login Complete…Almost There Fabric Zoning
FC
FC/FCoE Fabric
Initiator
Target
FCoE fabric zoning done the same as FC fabric zoning
Zoning is enforced at the FCF Zoning can be configured on the Nexus
5000 using the CLI or Fabric Manager If Nexus 5000 is an “FCoE Pass-
Through” device, zoning will be configured on the upstream core switch and pushed to the Nexus 5000
zone name EXAMPLE_Zone * fcid 0x10.00.01 [pwwn 10:00:00:00:c9:76:fd:31] [tnitiator] * fcid 0x11.00.01 [pwwn 50:06:01:61:3c:e0:1a:f6] [target] pwwn 10:00:00:00:c9:76:fd:31
pwwn 50:06:01:61:3c:e0:1a:f6
**Multi-hope FCoE with VE_Port not supported until Eaglehawk
FCF with Domain ID 10
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Host connected over unified wire to first hop access switch Access switch (Nexus 5000) is the
FCF Fibre Channel ports on the access
switch can be in NPV or Switch mode for native FC traffic
DCBX is used to negotiate the enhanced Ethernet capabilities
FIP is use to negotiate the FCoE capabilities as well as the host login process
FCoE runs from host to access switch FCF – native Ethernet and native FC break off at the access layer
FC
CNA
FC Fabric
ENode
Target
Ethernet Fabric
DCB capable Switch acting as an FCF
Unified Wire
Single Hop Design Today’s Solution
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Fabric A
CEE-DCBX
Generation 1 CNA
CIN-DCBX
Generation 2 CNA
Fabric B LAN Fabric
VN
VF Direct attach VN_Port to
VF_Port
Generation 1 CNA
limited to direct attached CNAs at the access
Utilized Cisco, Intel, Nuova Data Center Bridging Exchange protocol (CIN-DCBX)
Generation 2 CNA
Utilizes Converged Enhanced Ethernet Data Center Bridging Exchange protocol (CEE-DCBX)
Utilizes FCoE Initialization Protocol (FIP) as defined by the T.11 FC-BB-5 specification
Supports both direct and multi-hop attachment (through a Nexus 4000 FIP Snooping Bridge)
Single Hop Design The CNA Point of View
Nexus 5000 FCF-A
Nexus 5000 FCF-A
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Unified Fabric with FCoE CNA: Converged Network Adapter
• Standard drivers
• Same management • Operating System sees:
Dual port 10 Gigabit Ethernet adapter
Dual Port 4 Gbps Fibre Channel HBAs
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
vfc2
Eth1/2 PC1
vfc1
vfc3
mac-address
The Virtual Fibre Channel Interface Binding the vfc
Virtual Fibre Channel Interface (vfc) : Where is the logical Fibre Channel wire is terminated (the FCF)
Today this corresponds to an “F_Port”
Three options for binding a vfc interface : Physical Interface: Direct Attach CNAs and FCoE_NPV devices (future)
Single link port-channel: Direct Attach CNAs connected via a two port vPC
MAC-Address over an Ethernet Cloud: through a FIP-Snooping Device
Nexus 5000 FCF
Nexus 4000 FIP-Snooping
Nexus 5000 FCF
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
VLAN 10,30
VLAN 10,20
FCoE VLANs are treated differently than native Ethernet VLANs
No flooding, MAC learning, broadcasts, etc.
The FCoE VLAN must not be configured as a native VLAN
FIP uses native VLAN
FCoE VLANs should not be configured on Ethernet links that are not carrying FCoE traffic
Unified Wires must be configured as trunk ports and STP edge ports
! VLAN 20 is dedicated for VSAN 2 FCoE traffic (config)# vlan 20 (config-vlan)# fcoe vsan 2
VSAN 2
STP Edge Trunk
Fabric A Fabric B LAN Fabric
Nexus 5000 FCF
Nexus 5000 FCF
VSAN 3
Single Hop Design The FCoE VLAN
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
VLAN 10,30
VLAN 10,20
FCoE Fabric ‘A’ will have a different VLAN topology than FCoE Fabric ‘B’ which are different from the LAN Fabric
PVST+ allows unique topology per VLAN
MST requires that all switches in the same Region have the same mapping of VLANs to instances
MST does not require that all VLANs be defined in all switches
A separate instance must be used for FCoE VLANs
Recommended: three separate instances – native Ethernet VLANs, SAN ‘A’ VLANs and SAN ‘B’ VLANs
spanning-tree mst configuration name FCoE-Fabric revision 5 instance 5 vlan 1-19,40-3967,4048-4093 instance 10 vlan 20-29 instance 15 vlan 30-39
Fabric A Fabric B LAN Fabric
VSAN 3 VSAN 2
VLAN 10
Nexus 5000 FCF-A
Nexus 5000 FCF-B
Single Hop Design The FCoE VLAN and STP
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• VSANs use VLAN hardware table resources
• FCoE requires a VLAN and a VSAN that you bind the VLAN to. • Hence for each FCoE VSAN you should count using 2 VLANs
• Enabling FCoE burns two internal VSAN/VLAN resources • vFC binds to the Port-Channel, as long as there is one single port in
the port-channel attached to the switch
VLANs and VSANs FCoE Considerations
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• Optimal layer 2 LAN design often leverages Multi-Chassis Etherchannel (MCEC)
• Nexus utilizes Virtual Port-Channel (vPC) to enable MCEC either between switches or to direct attached servers (using LACP or static port-channels)
• MCEC provides network based load sharing and redundancy without introducing layer 2 loops in the topology
• MCEC maintains the separation of LAN and SAN high availability topologies
FC maintains separate SAN ‘A’ and SAN ‘B’ topologies
LAN utilizes a single logical topology Direct Attach vPC Topology
MCEC
vPC Peers
vPC Peer Link
Fabric A Fabric B LAN Fabric
Nexus 5000 FCF-A
Nexus 5000 FCF-B
Single Hop Design What is MCEC??
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• vPC enabled topologies with FCoE must follow specific design and forwarding rules…
• A ‘vfc’ interface can only be associated with a single-port port-channel
• While the port-channel configurations are the same on N5K-1 and N5K-2, the FCoE VLANs are different
• vPC configuration works with Gen-2 FIP enabled CNAs ONLY
• FCoE VLANs are ‘not’ carried on the vPC peer-link
• FCoE and FIP ethertypes are ‘not’ forwarded over the vPC peer link
Direct Attach vPC Topology
VLAN 10,30
VLAN 10,20 STP Edge Trunk
VLAN 10 ONLY HERE!
Fabric A Fabric B LAN Fabric
Nexus 5000 FCF-A
Nexus 5000 FCF-B
Single Hop Design Unified Wires and MCEC
vPC contains only 2 X 10GE links – one to each Nexus 5000
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• Multi-hop FCoE networks allow for FCoE traffic to extend past the access layer (first hop)
• In Multi-hop FCoE the role of a transit Ethernet bridge needs to be evaluated
Avoid Domain ID exhaustion
Ease management
• FIP Snooping is a minimum requirement suggested in FC-BB-5
• Fibre Channel over Ethernet NPV (FCoE-NPV) is a new capability intended to solve a number of design and management challenges
FCF
FCF
SAN B SAN A
DCB Capable Ethernet Switch
DCB Capable Ethernet Switch
VN
VN
VF
VF
Multi - Hop Design FCoE Pass-Through Options
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121 128
CLI Configuration Sample
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
Sample Working Topology
NPV Core
NPV Edge
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
NPV Edge Switch: pod7-5020-51(config)# feature fcoe
FC license checked out successfully
fc_plugin extracted successfully
FC plugin loaded successfully
FCoE manager enabled successfully
FC enabled on all modules successfully
pod7-5020-51(config)# feature npv
Verify that boot variables are set and the changes are saved. Changing to npv mode erases the current
configuration and reboots the switch in npv mode. Do you want to continue? (y/n):y
NPV Core Switch: pod3-9216i-70(config)# feature npiv
pod3-9216i-70(config)# feature fport-channel-trunk
Admin trunk mode has been set to off for
1- Interfaces with admin switchport mode F,FL,FX,SD,ST in admin down state
2- Interfaces with operational switchport mode F,FL,SD,ST.
Enabling the Required Features
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• NPV Edge Switch:
pod7-5020-51(config)# vsan database
pod7-5020-51(config-vsan-db)# vsan 10
• NPV Core Switch:
pod3-9216i-70(config)# vsan database
pod3-9216i-70(config-vsan-db)# vsan 10
pod3-9216i-70(config-vsan-db)# vsan 10 interface fc1/12
Configure the VSANs
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• NPV Core Switch:
pod3-9216i-70(config)# interface port-channel 1
pod3-9216i-70(config-if)# switchport mode f
pod3-9216i-70(config-if)# switchport trunk mode on
pod3-9216i-70(config-if)# channel mode active
pod3-9216i-70(config-if)# interface fc2/13, fc2/19
pod3-9216i-70(config-if)# switchport mode f
pod3-9216i-70(config-if)# switchport rate-mode dedicated
pod3-9216i-70(config-if)# switchport trunk mode on
pod3-9216i-70(config-if)# channel-group 100 force
fc2/13 fc2/19 added to port-channel 100 and disabled
please do the same operation on the switch at the other end of the port-channel, then do "no shutdown" at both end to bring them up
pod3-9216i-70(config-if)# no shut
Configure Trunking F_Port Port Channel
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• NPV Edge Switch:
pod7-5020-51(config)# interface san-port-channel 1
pod7-5020-51(config-if)# switchport mode np
pod7-5020-51(config-if)# switchport trunk mode on
pod7-5020-51(config-if)# interface fc2/1-2
pod7-5020-51(config-if)# switchport mode np
pod7-5020-51(config-if)# switchport trunk mode on
pod7-5020-51(config-if)# channel-group 1
fc2/1 fc2/2 added to port-channel 1 and disabled
please do the same operation on the switch at the other end of the port-channel, then do "no shutdown" at
both ends to bring it up
pod7-5020-51(config-if)# no shut
Configure Trunking F_Port Port Channel
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
pod7-5020-51(config)# vlan 10 pod7-5020-51(config-vlan)# fcoe vsan 10
pod7-5020-51(config-vlan)# interface ethernet 1/3 pod7-5020-51(config-if)# switchport mode trunk pod7-5020-51(config-if)# switchport trunk allowed vlan 1,10 pod7-5020-51(config-if)# spanning-tree port type edge trunk
Warning: Edge port type (portfast) should only be enabled on ports connected to a single host. Connecting hubs, concentrators, switches, bridges, etc... to this interface when edge port type (portfast) is enabled, can cause temporary bridging loops. Use with CAUTION
pod7-5020-51(config-if)# interface vfc3 pod7-5020-51(config-if)# bind interface ethernet 1/3 pod7-5020-51(config-if)# vsan database pod7-5020-51(config-vsan-db)# vsan 10 interface vfc3 pod7-5020-51(config-vsan-db)# interface vfc3 pod7-5020-51(config-if)# no shut
Configure FCoE on NPV Edge Switch
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
pod7-5020-51# sh npv flogi-table -------------------------------------------------------------------------------- SERVER EXTERNAL INTERFACE VSAN FCID PORT NAME NODE NAME INTERFACE -------------------------------------------------------------------------------- vfc3 10 0x0f0100 21:00:00:c0:dd:12:04:f3 20:00:00:c0:dd:12:04:f3 Spo1
Total number of flogi = 1.
Verify that the VFC interface is pinned to the SAN Port Channel
pod7-5020-51# sh npv traffic-usage
NPV Traffic Usage Information: ---------------------------------------- Server-If External-If
---------------------------------------- vfc3 san-port-channel 1 ----------------------------------------
Verify NPV Fabric Connectivity (Edge)
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
pod3-9216i-70(config)# show flogi database
--------------------------------------------------------------------------------
INTERFACE VSAN FCID PORT NAME NODE NAME
--------------------------------------------------------------------------------
fc1/12 10 0x0f00dc 21:00:00:20:37:a9:cd:6e 20:00:00:20:37:a9:cd:6e
fc1/12 10 0x0f00e0 21:00:00:20:37:a9:89:7e 20:00:00:20:37:a9:89:7e
fc1/12 10 0x0f00e2 21:00:00:20:37:af:de:85 20:00:00:20:37:af:de:85
fc1/12 10 0x0f00e4 21:00:00:20:37:a9:d6:49 20:00:00:20:37:a9:d6:49
fc1/12 10 0x0f00e8 21:00:00:20:37:a9:d7:bf 20:00:00:20:37:a9:d7:bf
fc1/12 10 0x0f00ef 21:00:00:20:37:a9:94:89 20:00:00:20:37:a9:94:89
port-channel 1 1 0x670102 24:01:00:0d:ec:a3:da:40 20:01:00:0d:ec:a3:da:41
port-channel 1 10 0x0f0100 21:00:00:c0:dd:12:04:f3 20:00:00:c0:dd:12:04:f3
Total number of flogi = 8.
Verify on Core that N5K and FCoE Workstation Are Logged into the Fabric.
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public BRKSAN-1121
• NPV Core Switch:
pod3-9216i-70(config)# zone name npv_vsan10 vsan 10 pod3-9216i-70(config-zone)# member pwwn 21:00:00:20:37:a9:cd:6e pod3-9216i-70(config-zone)# member pwwn 21:00:00:20:37:a9:89:7e pod3-9216i-70(config-zone)# member pwwn 21:00:00:20:37:af:de:85 pod3-9216i-70(config-zone)# member pwwn 21:00:00:c0:dd:12:04:f3 pod3-9216i-70(config-zone)# exit pod3-9216i-70(config)# zoneset name npv_v10_zs vsan 10 pod3-9216i-70(config-zoneset)# member npv_vsan10 pod3-9216i-70(config-zoneset)# zoneset activate name npv_v10_zs vsan 10 Zoneset activation initiated. check zone status
pod3-9216i-70(config)# show zoneset active zoneset name npv_v10_zs vsan 10 zone name npv_vsan10 vsan 10 * fcid 0x0f00dc [pwwn 21:00:00:20:37:a9:cd:6e] * fcid 0x0f00e0 [pwwn 21:00:00:20:37:a9:89:7e] * fcid 0x0f00e2 [pwwn 21:00:00:20:37:af:de:85] * fcid 0x0f0100 [pwwn 21:00:00:c0:dd:12:04:f3]
Configure Zoning