IBM Spectrum Virtualize - BP Zoning 101-V9.2

IBM Spectrum Virtualize Best Practice Zoning 101
Chuck Laing
Senior Technical Staff Member
IBM GTS SO/IS Storage Delivery
© Copyright IBM Corporation 2016. Technical University/Symposia materials may not be reproduced in whole or in part without the prior written permission of IBM.
Session objectives
• By the end of this session you should know:

• Zoning is one of the most important configurations for proper operation of:
1. Redundancy
2. Failover Resiliency
3. Throughput/Performance and proper isolation
• Basic “ground rules and achieving BP rules using:

• SVC Host definitions
• Mapping
• Masking
• Zoning
• The top 10 IBM Spectrum Virtualize (SVC) Best Practices (BP) zoning tips
• How to follow the BP rules of configuration to achieve :
• Proper dual core zoning for back-end devices
• Proper multiport (4 of more) for hosts and clusters
• Proper zoning for AIX LPM configurations
• Proper zoning for Enhanced Stretch Cluster
© Copyright IBM Corporation 2016. Technical University/Symposia materials 1

IBM Systems Technical Events | ibm.com/training/events may not be reproduced in whole or in part without the prior written permission of
IBM.
BASIC BEST PRACTICE GROUND RULES
IBM Spectrum Virtualize Zoning - Best Practices 101

Define port layout usage,
Establish masking and mapping to 4 or more host port consistencies

IBM.
SVC Port layout using Fully populated 4 port HBA @16Gb speed
GTS Best Practice Policy:
• Use and configure a minimum of 12 and optionally 16 port nodes with
fully populated 4 Port HBA at 16Gbs capability
Reason for Best Practice:

1. To ensure maximum bandwidth and optimal configurations for specific ports to
specific uses and aligning the internal allocation of hardware CPU cores and
software I/O threads to those ports.
2. To align GTS with STG and Redbook published recommendations
Action/s:
• The best practices will leverage Hursley supported configurations and tested
environments
– Varying from these recommendations may result in unexpected consequences. In addition, configuring as recommended
above will ensure the ability to replace nodes non-disruptively in the future.
– Do not zone the same ports:
• Host/Storage ports to ports designated for Inter-node use or Replication use in the 8/12/16 port configurations
• Inter-node and replication ports.
This is to minimize any B2B credit exhaustion situations, due to long distance latencies introduced by replication from
tying up buffers needed by hosts, storage or inter-node communications.
– With 12 or more ports per node we highly recommend 4 ports to be dedicated for inter-node traffic, especially when high
write data rates are expected (3Gb and greater) as all writes are mirrored over these ports within an I/O Group.
– GTS does not support dedicating certain ports specifically for hosts and other ports for storage as it negates the
full duplex capability of the fibre channel HBA ports.
IBM.
DH8 Port Layout 7.6.x– 16 Gb Capable speeds for New Builds
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 1213 14 15 16 1 2 34 5 6 7 8 9 10 11 1213 14 15 16
port with
number Slot 1 Slot 2 Slot 3 Slot 5 Slot 1 Slot 2 Slot 3 Slot 5
wwpn #
16G HBA Node 1 Node 2 embedded
8G HBA I/O Group 0 I/O Group 0
DH8 1 Building Block – 16 FC ports per node

IBM.
SVC Port layout using mixed speed port scenarios
GTS Best Practice Policy:
• Mixed FC port speeds in new and existing SVC environments* should be
avoided whenever possible
Reason for Best Practice:

• To ensure maximum bandwidth appropriately for port device attachments.
Action/s:
• The best practices will leverage Hursley supported configurations and tested
environments
– Recommendation - no mixed FC port speeds whenever possible - either all 8Gbps or all 16Gbps … else
– Use of mixed FC port speeds* should be limited to configurations with same speed ports dedicated (through zoning) to
either host/storage or inter-node/replication: Do not use mixed speeds for the same purpose (single host multipath with
mixed speeds is officially unsupported configuration).
– Same rules apply to another mixed port speeds configuration at 4x 8Gbps and 4x 16Gbps, again with dedicated ports
utilization enforced by zoning
– Mixed FC-iSCSI configurations should limit iSCSI ports oversubscription, to avoid turning SVC into a "NAS-like" appliance
– use the same as FC ports.
– For existing account environments where 2 port HBA cards at 16Gbs ports are used, you should ensure that the 8Gbs
HBAs remain in slots 1 and 2
– (*) One such configuration is shown is the next slide

IBM.
DH8 Port Layout 7.5.x– 8 & 16 Gb Capable Combination
- for Existing Builds
• Could have up to 14 ports,
– 4 port 8Gb HBA in slots 1,2,5 giving you 12 ports
– 2 port 16Gb card in slot
Physical 1 2 1 2 2 2 2 2 3 3 5 5 1 2 1 2 2 2 2 2 3 3 5 5
port 1 2 2 2 1 2 3 4 1 2 1 3 1 2 2 2 1 2 3 4 1 2 1 3 Logical
number port with
1 2 34 56 78 9 10 11 12 1 2 34 56 78 9 10 11 12 wwpn #
Slot 1 Slot 2 Slot 3 Slot 5 Slot 1 Slot 2 Slot 3 Slot 5 embedded
16G HBA
Node 1 Node 2
8G HBA
I/O Group 0

IBM.
Definitions and Rules
Calculating Port to Port Logins, not to exceed 16
• Regarding the calculation of the message login limit, here is how it works:
• This limit applies to any single node X, as seen by another node Y.
• Only 16 port logins are allowed from one node to any other node in an SAN fabric
• New error code 888, alert 801 will be generated if 16 exceeded
• Must correct zoning or masking or a combination of both zoning and masking to clear this issue
• To check whether the limit is exceeded, count the number of distinct ways by which a port on node X
can log into a port on node Y. This must not exceed 16.
• Only ports that are permitted to carry cluster message traffic need be included in the count.
• Any given two Nodes can see each other for messaging in 8 different ways (4 on each switch)
• Only the ports that can "see" each other (based upon switch zoning) need be included.
• Note : Port logins associated with host/storage ports, strictly speaking,
can be disregarded since they don't carry message traffic. For separate
reasons however, Hursley development recommends that the number of
Spectrum Virtualize ports in any given single zone be limited to avoid the
overheads associated with a large numbers of logins.

IBM
7 Systems Technical Events | ibm.com/training/events may not be reproduced in whole or in part without the prior written permission of
IBM.
Definitions and Rules
Three (3) types of Port Masking
• There are three(3) types of Port masking to be aware of with the new
code and DH8
• Host port masking (mkhost)
• Default port masking, with 64 bits, used to be "1111" now it is "...111111111111". reading from right to
left...1 for each of the 12 ports. Leaving the default of 1 enables the flow of IO to pass through the node
port, "0" means to block IO on the port.
• Benefits
• zoning changes
• Less complexity for Vdisk IOgrp moves,
• Changes introduced in the SAN, or the hosts, with less complexity
• The MM/GM port mask (partner_fc_mask)

• Allows only MM/GM replication traffic to flow through the enable ports,
• Note: Other traffic can flow here from other connected devices, but defeats the purpose of dedicating MM/GM traffic .
We dedicate and use port 6&12 for MM/GM (partner_fc_port_mask) so the mask would be 100000100000
• The Intra-cluster communication port mask (local_fc_port_mask)
• Also dedicated at 000001000010 for 12 ports meaning ports 2, & 7 are enabled with 1's and the rest are
blocked.
• Bottom-line: We have so much field experience behind these best practices, that this is what we
recommend. The most common reoccurring issue is when LUNs are moved around between IOgrps or
shared to clusters, the zoning gets redone, but most SME's forget about or don't understand the port
mask and so don't change it and then the redundancy is lost when change is made, resulting in IO
blocked and outages...this we made it a best practice to reduce the complexity of managing the LUN to
host mapping.
IBM
IBM.
Host Zones and Vdisk Mapping
Spectrum Virtualize Preferred Node Scheme
• Hosts with four (or more) Host Bus Adapters (HBAs)

• Takes a little more planning.
• Because eight paths are not an optimum number, you must instead configure your Spectrum
Virtualize Host Definitions (and zoning) as though the single host is two or more separate hosts
• During Vdisk assignment, alternate which Vdisk is assigned to one of the “pseudo-hosts”, in a
round robin fashion (a pseudo-host is nothing more than another regular host definition in the
Spectrum Virtualize host config. Each pseudo-host will contain 2 unique host WWPNs, 1 WWPN
mapped to each fabric)
• Note: A pseudo-host, is not a defined function or feature of the Spectrum Virtualize. If you need to define a
pseudo-host, you are simply adding another host id to the Spectrum Virtualize host config. Instead of creating
one host id with 4 WWPNs, you would define 2 hosts with 2 WWPNs. This is now the reference for the term
pseudo-host.
• Note: Be careful not to share the Vdisk to more than two adapters per host, so as to not oversubscribe the
number of datapaths per vdisk per host

IBM.
IBM Spectrum Virtualize BP Rules
• The top10 Spectrum Virtulize (SVC) zoning tips:

1. Do not separate Host and Storage Ports - use together for full duplex bandwidth
2. One host initiator port to multiple SVC node target ports
3. 8 paths per Vdisk max limitations …BUT recommend BP of – no more than 4 – no less than 4
4. Two ports per node usage for Replication - Metro Mirror/Global Mirror (MM/GM)
5. 2 ports minimum to 4 ports maximum for Inter-node (messaging) communication, no ISLs
6. One Zone per SVC Node per Fabric for each back-end storage device
7. Isolate Storage Frames/Units to separate zones – no massive multi-storage unit zones
8. The same back-end ports must be seen across all Nodes
9. Zone for growth across Core – Edge designs to minimize IO traversing across ISLs
10. Isolate Hosts to separate zones – including multiport (4 or more ports) / host/s

IBM.
BASIC BEST PRACTICE RULE # 1

Do not separate Host and Storage Ports - use together for full duplex bandwidth

IBM.
Do not separate Host and Storage Ports
• Use together for full duplex bandwidth
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical
1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
port Logical
number port with
1 2 3 4 5 6 7 8 9 10 11 1213 14 15 16 1 2 34 5 6 7 8 9 10 11 1213 14 15 16
Slot 3 Slot 3 wwpn #
Slot 1 Slot 2 Slot 5 Slot 1 Slot 2 Slot 5
embedded
16G HBA Node 1 Node 2
8G HBA
I/O Group 0 I/O Group 0

IBM.
New Storage Zoning Schema per Iogrp 16 Port Node
Do not separate port usage, Use the

Production same ports for full duplex bandwidth Production
SAN Fabric SAN Fabric
A B
STG
Zone-3
STG Host
Zone-1 Zone-2
Host
Zone-4
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.

One host initiator port to multiple SVC node target ports

IBM.
Host Zones – 1 host initiator to 2 SVC node target ports
• Should be (1:2) zoning Host

A1 B1
1 host initiator port
Fabric1 Edge1 Fabric2 Edge1
2 SVC Node
Target ports
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.
Host Zones – Too few
• No (1:1) zoning Host

A1 B1
1 host initiator port
1 SVC Node
Target ports
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.

8 paths per Vdisk max limitations …BUT recommend BP of:
- No more than 4
- No less than 4

IBM.
Host Zones – 4 paths per Vdisk
• 2 preferred – 1 red & 1 blue Host

• 2 non preferred – red & blue A1 B1
Preferred Paths Non Preferred Paths
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.
Host Zones – too many paths and target ports
Host
A1 B1
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.
Host Zones – Too few
• Spectrum Virtualize - DH8 Host

A1 B1
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.
Right or Wrong?
Fabric_A Zone Definitions Fabric_B Zone Definitions

A1=10:00:00:00:c9:3f:75:37 b03vio210 B1=10:00:00:00:c9:3f:74:16
A2=10:00:00:00:c9:3f:80:54 A1 A2 A3 A4 B1 B2 B3 B4 B2=10:00:00:00:c9:3f:f7:2f
A3=10:00:00:00:c9:42:89:ac B3=10:00:00:00:c9:42:b4:4b
A4=10:00:00:00:c9:42:80:16 B4=10:00:00:00:c9:42:b6:18
DIR1 SAN Fabric DIR2 SAN Fabric
In this example we show

SVC Host Definitions 1 SVC Host definitions and
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P4 id:8
name:b03vio210_allvg
2 Fabric zones, per host per Fabric.
HBA 1 HBA 2 HBA 1 HBA 2 HBA 1 HBA 2 HBA 1 HBA 2
10000000C942B618
Node 1 Node 2 Node 3 Node 4 10000000C942B44B This does not allow proper failover from
10000000C9428016
iogrp 0 iogrp 1 10000000C94289AC Primary to Alternate nodes within the
2048 LUNs max 2048 LUNs max
10000000C93FF72F SVC iogrp.
10000000C93F7416
wwpn 4 Node SVC MAX Vdisk 4096
10000000C93F8054
10000000C93F7537 See the next page for proper zoning
5005076801

IBM.
Right
b03vio210_1 b03vio210_2 b03vio210_3 b03vio210_4
A1 B1 A2 B2 A3 B3 A4 B4
Fabric_A Zone Definitions Fabric_A Zone Definitions
Fabric_B Zone Definitions A4=10:00:00:00:c9:42:80:16
Fabric_B Zone Definitions A3=10:00:00:00:c9:42:89:ac
B1=10:00:00:00:c9:3f:74:16
Fabric_A Zone Definitions B2=10:00:00:00:c9:3f:f7:2f
A1=10:00:00:00:c9:3f:75:37
Fabric_B Zone Definitions
Fabric_B Zone Definitions
Fabric_A Zone Definitions B4=10:00:00:00:c9:42:b6:18
B3=10:00:00:00:c9:42:b4:4b
A2=10:00:00:00:c9:3f:80:54
DIR1 SAN Fabric DIR2 SAN Fabric
SVC Host Definitions

id:1 name:b03vio210_1 In this example we show
10000000C93F7537 4 seperate SVC Host definitions and
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P4 10000000C9428016
2 seperate Fabric zones, per host
HBA 1 HBA 2 HBA 1 HBA 2 HBA 1 HBA 2 HBA 1 HBA 2
id:2 name:b03vio210_2 per Fabric
10000000C93F8054
Node 1 Node 2 Node 3 Node 4 10000000C93FF72F
This does allow proper failover from
iogrp 0 iogrp 1 id:3 name:b03vio210_3 Primary to Alternate nodes within the
2048 LUNs max 2048 LUNs max 10000000C94289AC
10000000C942B44B SVC iogrp.
wwpn 4 Node SVC MAX Vdisk 4096
5005076801 id:4 name:b03vio210_4
10000000C93F7416
10000000C942B618

IBM.
Summary - 3 factors determine right/wrong paths
• Example of correct/incorrect number of datapaths with Host to Spectrum Virtualize

• Correct datapathing has 3 factors
• Proper zoning
• Proper Spectrum Virtualize Host definitions (Spectrum Virtualize logical config of the host def)
• Proper redundancy for the Spectrum Virtualize preferred /non preferred pathing

IBM.

Two ports per node usage for Replication - Metro Mirror/Global Mirror (MM/GM)

IBM.
Replication Zoning Schema - remotefcportmask 1001000000

I/O Group 0
Physical Node 1 Node 2
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 wwpn #
16G HBA
1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 embedded
8G HBA
WAN direct connection to the WAN direct connection to the

FCIP connection to the Node ports FCiP connection to the Node ports
GM/MM zone A GM/MM zone B
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.

2 ports minimum to 4 ports maximum for Inter-node (messaging) communication
No inter-node ISLs

IBM.
No ISL hops for Inter-node - localfcportask 110000110000
Inter-node Inter-node Production Production

Internal SAN Fabric
Zone Zone SAN Fabric
SAN Fabric A B
Dedicated Dedicated
B
C D
STG STG
STG Zone-2 Zone-3 STG
Zone-1 Zone-4
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM
IBM.

One Zone per SVC Node per Fabric for each back-end storage device

IBM.
Making 1 zone per Node per Fabric with

Production the same ports from a single backend Production
SAN Fabric SAN Fabric
A
storage unit, will reduce the overheads
B
associated with large numbers of logins
STG
Zone-2
STG
Zone-3
STG STG
Zone-1 Zone-4
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.
This shows only 2 zones Back-end Storage

Spanning multiple Nodes HBA1 HBA2
P P P P
1 2 1 2
SAN Fabric 1 SAN Fabric 2

IBM.

Isolate Storage Frames/Units to separate zones – no massive multi-storage unit
zones

IBM.
Correct Dual Core Zoning to Multiple Storage devices
• Create separate zones for each Storage device
V7000
DS8K-1 DS8K-2
CAUTION 1 2 1 2 3 4 1 2 3 4 CAUTION
XIV
3 4 1 2
1 2 1 2
3 4 3 4
Disconnect all Disconnect all
supply power for supply power for
complete isolation 1 2 complete isolation
1 2
2 1
2 1
complete isolation complete isolation
4 3 4 3
2 1 2 1
2 1 4 3
4 3 2 1 4 3 2 1 2 1
CAUTION CAUTION
V7K XIV DS8K-1 DS8K-2 V7K XIV DS8K-1 DS8K-2

Zones
Fabric1 Core1 Fabric 2 Core2
Zones Zones Zones Zones Zones Zones Zones

IBM.
Correct Dual Core Zoning to Multiple Storage devices
• Zone each device to only one core to avoid extra “ISL IO traversing” cutting down ISL hops
• Spread Multiple Storage devices as evenly possible across Cores
V7000
DS8K-1 DS8K-2
CAUTION 1 2 1 2 3 4 1 2 3 4 CAUTION
XIV
3 4 1 2
1 2 1 2
3 4 3 4
complete isolation 1 2 complete isolation
1 2
2 1
2 1
complete isolation complete isolation
4 3 4 3
2 1 2 1
2 1 4 3
4 3 2 1 4 3 2 1 2 1
CAUTION CAUTION
Storage Storage
Zone-1
Fabric1 Core1 Fabric 2 Core2
Zone-2

IBM.

The Same back-end ports must be seen across all nodes

IBM.
The same back-end ports are used for each Node Zone
XIV Storage
Module1 Module2 Module3 Module4 Module5 Module6
HBA1 HBA2 HBA1 HBA2 HBA1 HBA2 HBA1 HBA2 HBA1 HBA2 HBA1 HBA2
P P P P P P P P P P P P P P P P P P P P P P P P
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
STG STG STG STG

SAN
Zone1 Zone3 Fabric 1 SAN
Zone2 Zone4 Fabric 2
Up to 16 zones for an 8 node Cluster

2 zones /fabric/ node
Up to 16 Ports max from any
back-end storage

IBM.
The same back-end ports must be seen across all Nodes
XIV Storage
Module1 Module2 Module3 Module4 Module5 Module6
HBA1 HBA2 HBA1 HBA2 HBA1 HBA2 HBA1 HBA2 HBA1 HBA2 HBA1 HBA2
P P P P P P P P P P P P P P P P P P P P P P P P
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Same SAN
SAN 1
SAN 2
SAN 1
IBM Systems Technical Events | ibm.com/training/events

SAN 2
© Copyright IBM Corporation 2016. Technical University/Symposia materials
may not be reproduced in whole or in part without the prior written permission of
36
IBM.

Topology and zoning for growth across Core – Edge designs to minimize IO
traversing across ISLs

IBM.
General Spectrum Virtualize Best Practice Zoning Concepts
• Dual Core Fabrics require a breakout in cluster zoning to separate

Spectrum Virtualize node ports on core1 from core2
– Prevents/Minimizes IO from spanning/routing across ISLs and Edge switches in a Core/Edge
design
• Spectrum Virtualize node ports plugged into dir1 core1 should reside in a separate zone from
Spectrum Virtualize node ports plugged into dir1 core2.
• For MM/GM zones, take advantage of spreading the IO across both cores through careful
zoning
• Ensure the Intra-cluster/heartbeat zone is also isolated properly between each fabrics dual
cores (don’t span heartbeat across cores in one fabric)
– Note: - Never make zoning changes on redundant Fabrics at the same time
• Make changes on one fabric and wait 30 min in-between
• Please see the following link for zoning BP’s
– Cisco:
http://g25aciwas09.con.can.ibm.com:9080/ram/assetDetail/generalDetails.faces?guid=5EB2DDF9-7568-368A-7F99-
1BCFFF85E472
– Brocade:
http://g25aciwas09.con.can.ibm.com:9080/ram/assetDetail/generalDetails.faces?guid=57A9649D-A04B-5931-05DA-
F389F697D4D3

IBM.
Storage Zone Type – Good Core Edge Design – 1ISL Hops
Back-end Storage
Host X
HBA1 HBA2
A1 B1
P P P P
1 2 1 2
STG Host Host
STG Zone-2 Zone-1
Zone-1 Zone-2
IO request 1 ISL Hop1
Fabric1 Core1 Fabric 1 Edge 1 Fabric2 Edge1 Fabric 2 Core 1
1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5 1 2 1 2 2 2 2 2 3 3 3 3 5 5 5 5
Physical 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2 2 1 2 3 4 1 2 3 4 1 2 3 4
Logical
port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16
port with
wwpn #
8G HBA
I/O Group 0

IBM.
Dual Core Zoning/Cabling to Multiple Storage devices
• Zone each device to only one core to avoid extra “ISL IO traversing” cutting down ISL hops
• Round Robin / Spread Multiple Storage devices as evenly possible across Cores
DS8K_1 DS8K_2 XIV V7000
Fabric1 Core1 Fabric 1 Core2 Fabric2 Core1 Fabric 2 Core2

IBM.
Incorrect “Dual Core” fabric zoning example
- for a 2 node Spectrum Virtualize Cluster

IBM.
Slide provided by John Locke
Correct “Dual Core” fabric zoning example
-for a 2 node Spectrum Virtualize Cluster

IBM.
Slide provided by John Locke

Isolate Hosts to separate zones – including multiport (4 or more ports) / host/s

IBM.
Multiple Host - ESX Cluster to Spectrum Virtualize
port zoning
• Break down
Host_1a Host_1b Host_2a
Host_ Host_2b
A1
A1 B1 A1 B1 A1 B1 A1 B1
Fabric A Fabric B

IBM.
port zoning
• Break down
Host_ Host_2b
A1
A1 B1 A1 B1 B1 A1 B1
Fabric A Fabric B

IBM.
port zoning
• Break down
Host_ Host_2b
A1
A1 B1 A1 B1 B1 A1 B1
Fabric A Fabric B

IBM.
port zoning
• Break down
Host_1a Host_1b Host_2a Host_2b

A1
A1 B1 A1 B1 B1 A1 B1
Fabric A Fabric B

IBM.
Zoning Multi HBA hosts for Resiliency
• Right way to zone a multiport Host

• Make four zones (Red, Blue, Orange and Green zones)
• Spread IO across multiple Ports SVC ports
Multi Port Host

HBA1 HBA2
P P P P P
1 2 1 2 1 2
SAN Fabric 1 SAN Fabric 2

IBM.
Host Zones
Best Practices Concepts
• Planning and design for AIX Live Partition Mobility (LPM)
• Create two separate and isolated zones on each fabric for each LPAR
• Do not put both the active and inactive LPAR wwpns in either the same zone or same Spectrum Virtualize host
definition
• Create Pseudo host Spectrum Virtualize host definitions containing only two virtual wwpns, one from
each fabric
• Follow the rules shown in the next few slides
• Map LUNs to the virtual host FC HBA port wwpns not the physical host FCA Adapter wwpn
• Exception to map only boot from SAN LUNs to the physical host FC HBA wwpn
• Round Robin the Vdisks to the LPAR Pseudo and both the active and inactive LPARs to enable LPM
• Note: Following these rules will result in not having a degraded path error on the Spectrum Virtualize, and keep the 4 paths per Vdisk rule

IBM.
Dual VIOS to Multiple LPARs
Map LUNs within an LPAR in a round robin fashion to the active pseudo hosts. For example:
• Vdisk1 to Pseudo1 – Vdisk2 to Pseudo2 – Vdisk3 to Pseudo1 – Vdisk4 to Pseudo2
VIO Server1 P1 P5 VIO Server2
FCA1 FCA3
VP1.1aVFCA Active
P2 P6 Client
Pseudo1
Logical
P3 P7 VP7.1aVFCA Partition
FCA2 FCA4 (LPAR1)
VFCA1.1
P4
SAN VFCA5.1
P8 VP5.1aVFCA
Pseudo2
VP1.1i VP1.1a VP5.1a VFCA
VFCA2.1 VP5.1i VP3.1a
VFCA6.1
VP2.2i VP2.2a VP6.2a VP6.2a

VFCA3.1
Spectrum VP2.2aVFCA
VP3.1i Virtualize VFCA7.1
VP3.1a VP7.1a VP7.1i Pseudo1 Active
VFCA Client
VFCA4.1 VFCA8.1 VP8.2a Logical
VP4.2i VP4.2a VP8.2a VP8.2i Partition
VP6.2aVFCA
I (LPAR2)
Pseudo 2
VFCA
VP4.2a
Fame1 Hypervisor

IBM.
FCA1 FCA3 VP1.1aVFCA
P2 P6 Active
Pseudo1 Client
P3 P7 VP7.1aVFCA Logical
FCA2 FCA4
P4 SAN P8 VP5.1aVFCA Partition
(LPAR1)
Pseudo2
VP1.1i VP1.1a VP5.1a VP5.1i VP3.1aVFCA
SVC VP2.2aVFCA
VP3.1i VP3.1a VFCA4.1VP7.1i
VP7.1a Pseudo1 Active
VP8.2aVFCA Client
VP4.2i VP4.2a VP8.2a VP8.2i
I VFCA
VP6.2a Logical
Pseudo 2 Partition
(LPAR2)
VP4.2aVFCA
inactive & active LPM
vWWPN pairs
Fame1 Hypervisor
P1 P3
Avoiding this
P2 P4
. SAN VP1.1i VFCA Active
•VFCA
VFCA
LPM Pseudo1 Client
Logical
P3 … P64 VP7.1i VFCA
I Could go to Partition
Fame2 VP5.1i VFCA
(Pseudo
Frame2 or LPAR1b)
Hypervisor Pseudo2 During LPM
Frame3 VFCA
VP3.1i
P1 P3
P2 P4 During LPM the number of paths double from 4 to 8
. SAN
•VFCA
Starting with 8 paths per vdisk will render an unsupported 16
VFCA
P3 … P64 I paths during this time - could lead to IO interruption
Fame3
Hypervisor
IBM.
51
FCA1 FCA3 VP1.1aVFCA
P2 P6 Active
Pseudo1 Client
P3 P7 VP7.1aVFCA Logical
FCA2 FCA4
P4 SAN P8 VP5.1aVFCA Partition
(LPAR1)
Pseudo2
VP1.1i VP1.1a VP5.1a VP5.1i VP3.1aVFCA
SVC VP2.2aVFCA
VP3.1i VP3.1a VFCA4.1VP7.1i
VP7.1a Pseudo1 Active
VP8.2aVFCA Client
I VFCA
VP6.2a Logical
Pseudo 2 Partition
(LPAR2)
VP4.2aVFCA
inactive & active LPM
vWWPN pairs
Fame1 Hypervisor
P1 P3
P2 P4
. SAN VP1.1i VFCA Active
•VFCA
VFCA
LPM Pseudo1 Client
Logical
P3 … P64 VP7.1i VFCA
I Could go to Partition
Fame2 VP5.1i VFCA
(Pseudo
Frame2 or LPAR1b)
Hypervisor Pseudo2 During LPM
Frame3 VFCA
VP3.1i
P1 P3
P2 P4 During LPM the number of paths double from 4 to 8
. SAN
•VFCA
Starting with 8 paths per vdisk will render an unsupported 16
VFCA
P3 … P64 I paths during this time - could lead to IO interruption
Fame3
Hypervisor
IBM.
52
Is it resilient? - One VIOS Failure

FCA1 FCA3 VFCA
VP1.1a Active
P2 P6 Client
Pseudo1 Logical
VFCA Partition
P3 P7 VP7.1a (LPAR1)
FCA2 FCA4
SAN VFCA
VP5.1a
P4 P8 LPM
VFCA1.1 VFCA5.1 Pseudo2
VFCA
VP1.1i VP1.1a VP5.1a VP5.1i VP3.1a
VFCA2.1 VFCA6.1
VFCA3.1 VFCA
Spectrum VP2.2a
Virtualize VFCA7.1
VP3.1i VP3.1a Active
VP7.1a VP7.1i Pseudo1
VFCA4.1 Client
VFCA8.1 VFCA Logical
VP8.2a
Partition
(LPAR2)
VFCA
VP6.2aI
Pseudo 2
x
VFCA
VP4.2a
Fame1 Hypervisor
IBM.
Is it resilient? – One SAN Fabric Failure

FCA1 FCA3 VFCA
VP1.1a Active
P2 P6 Client
Pseudo1 Logical
VFCA Partition
P3 P7 VP7.1a (LPAR1)
FCA2 FCA4
SAN VFCA
VP5.1a
P4 P8
VFCA1.1 Pseudo2
LPM
VFCA5.1
VFCA
VP1.1i VP1.1a VP5.1a VP5.1i VP3.1a
VFCA2.1 VFCA6.1
VFCA3.1 VFCA
Spectrum VP2.2a
Virtualize VFCA7.1
VP3.1i VP3.1a Active
VP7.1a VP7.1i Pseudo1
Client
VFCA4.1 VFCA8.1 VFCA Logical
VP8.2a
Partition
(LPAR2)
VFCA
VP6.2aI
Pseudo 2
VFCA
VP4.2a x
Fame1 Hypervisor

IBM.
Spectrum Virtualize (SVC) ESC – Best Practice
Use Case example
Host Server Host Server
Clustered Servers
where applicable
Mission Critical vols

Node 1 Node 2
Example 1 Stretched I/O Group
SVC Cluster 3
SVC Cluster 0
SVC Global Mirror Node 1 Node 2
Site-1 Site-2 ASYNC
DR Site
• Quorum
Storage Pool 1 Storage Pool 2
Storage Pool 3
B COPY
C COPY
Mission Critical Block for Applications
HA Storage: can sustain .3-1 ms latency hit) DR, Dev, Test, QA .. etc.
IBM.
Spectrum Virtualize (SVC) ESC – Site-1 & Site-2 current building block example
Example “Use Case” Scenario description:

Public VSAN Public VSAN
• Nodes are stretched across sites to make an SVC
3X / 3X /
node node
Enhanced Stretch Cluster (ESC) for High Availability
(HA) application requirements
Private VSAN Private VSAN
SVC DH8/4nodes /
3HBAs @ 8Gb • A Core Public enterprise class Cisco VSAN is configured
Cisco MDS
9710 Core
Cisco MDS
9710 Core
for Host/Storage port connections
2X
2X
2X
• A Core Private Cisco VSAN is configured for Node to
2X
Node Communication port connections
2X 2X
Node to
node
Node to
node
• 12 Port Nodes are used with 4 ports each at 8Gb port
2X
2X speeds
• Edge Switches per port count requirements are used for

GM a true BP Core-Edge design to allow future growth
4X/Flash, 6X/XIV, 4X/V7K 4X/Flash, 6X/XIV, 4X/V7K
scalability
GM
4x ISL / Edge 4x ISL / Edge
• Storage building blocks of Flash, XIV and V7000 are used

for a 3 tier or mixed tier approach
9396 Edge x 3
9396 Edge x 3
• Separate Storage zones are created for each device
providing isolated zones from each device type
Host Servers
1X / host 1X / host • Host Servers are connected to the Edge Switches
• Edge Switches are ISLed to the Core Public VSANs
Servers • FCIP Routers are used for GM to the DR site. SVC GM

1X / SVC Node
1X / SVC Node
ports are plugged directly into the FCIP Routers

FCIP Router 9250i FCIP Router 9250i
IBM.
Spectrum Virtualize (SVC) ESC – DR site current building block example
16X (2x/ SVC Node)
Example “Use Case” Scenario

FCIP Router GM GM
9250i FCIP Router
9250i
description:
Cisco MDS
9710 Core Cisco MDS
Public
9710 Core • A Core Public enterprise class Cisco VSAN is
VSAN Public configured for Host/Storage port connections
3x / 3x / VSAN
Private nod nod
e Private
VSAN e VSAN • A Core Private Cisco VSAN is configured for Node
to Node Communication port connections
SVC DH8
2
X 2
2
X • 12 Port Nodes are used with 4 ports each at 8Gb
X 2
X
port speeds
2
2
X
X
2 2 • Edge Switches per port count requirements are
X X
2 2
used for a true BP Core-Edge design to allow future
X
2
X growth scalability
2
X
X
2 2
X
2
X
2 • Storage building blocks of Flash, XIV and V7000 are
X
X used for a 3 tier or mixed tier approach
4x/Flash, 4x/Flash, • Separate Storage zones are created for each device
6x/XIV, 6x/XIV,
2x/V7K 2x/V7K providing isolated zones from each device type
8x ISL / Edge
8x ISL / Edge
• Host Servers are connected to the Edge Switches
Quorum • Edge Switches are ISLed to the Core Public VSANs

9396 Edge x 3 9396 Edge x 3
??? Servers
1X / host 1X / host
• FCIP Routers are used for GM to the DR site. SVC
GM ports are plugged directly into the FCIP Routers

Servers may not be reproduced in whole or in part without the prior written permission of
IBM.
Spectrum Virtualize (SVC) ESC site-1 & site-2 synchronous with DR asynchronous
Putting it all together in the example template
2X @ 16Gb/s LW CWDM SFP
2X @ 16Gb/s LW CWDM SFP
Site1 Site2
FCIP GM
Priv Dark Fibre
Pub Dark Fibre
4X
4X
4X
4X
2X @ 16Gb/s CWDM LW SFP
2X @ 16Gb/s CWDM LW SFP
DR Site

IBM.
Session Evaluations
YOUR OPINION MATTERS!
1 2 3 4
Submit four or more session
evaluations by 5:30pm Wednesday
to be eligible for drawings!
*Winners will be notified Thursday morning. Prizes must be picked up at
registration desk, during operating hours, by the conclusion of the event.

IBM.
Continue growing your IBM skills
ibm.com/training
provides a comprehensive
portfolio of skills and career
accelerators that are designed
to meet all your training needs.
If you can’t find the training that is right for you with our
Global Training Providers, we can help.
Contact IBM Training at [email protected]
Global Skills Initiative

IBM.
Thank you!
For you interest
and attendance

IBM.
IBM SPECTRUM VIRTUALIZE
Backup Slides

IBM.
Most common zoning questions - BP recommendations
• What is the max number of WWNN/WWPNs devices?

– Always check the Max Limit configuration URL for the most current updates
• Search for the term "configuration limits and restrictions" in the Search support and downloads search box at the following
website:
– www.ibm.com/storage/support/2145
– Currently 1024 WWNN per cluster and 1024 WWPNs per cluster
• Recommendation - The more wwpns per wwnn, the more throughput up to 16
• How many IO connections/zones per storage device to the Spectrum Virtualize

should I zone?
– 16 from any “one” storage device unit zoned with all Spectrum Virtualize node ports
• All Spectrum Virtualize nodes must see same set of LUNs from disk
controller
– Otherwise degraded mode on controller and/or MDisks

IBM.
• Should I mask host connections through zoning or Spectrum Virtualize

configurations?
– Best practice is to use the default value of …..111111111111 (all ports enabled) and control masking
through zoning
– Use host type Target Port Group (tpgs) for Solaris hosts, (hpux) for HP/UX, (openvms) for HP
AlphaServer/HP Intergrity and (generic) for everything else
– Separate disk and tape IO on host HBAs
• What works better, 4 or 8 paths per Vdisk?
– Recommendation is 4 paths per Vdisk

IBM.
• How many Iogrps should I map to a host? 4?

– Recommendation is to size per throughput and number of hosts per cluster
– May need to isolate/sement host activity/chatter from one Iogrp to another
– For more horsepower – spread IO across more LUNs from more Iogrps
• Can a host have only one connection to the Spectrum Virtualize ?
– Dual host HBA connections

IBM.
Disk Controller Best Practices
• Maximum of 1024 WWNNs
– EMC DMX/SYMM, All HDS and SUN/HP HDS clones use one WWNN per port; each appears as a
separate controller to Spectrum Virtualize
• Map LUNs through up to 16 FA ports

– Results in 16 WWNNs/WWPNs used out of the max of 1024
– IBM, EMC Clariion, HP, etc. use one WWNN per subsystem; each appears as a single controller with
multiple ports/WWPNs
• Maximum of 16 ports/WWPNs per WWNN using 1 out of the

max of 1024
• Greater Qdepth is archived with fewer Mdisks

IBM.
Disk Controller Best Practices
• DS4K/5K – EMC Clariion/CX
– Both have preferred controller architecture
• Spectrum Virtualize honors this configuration
– Use minimum of 4 and preferably 8 ports or more up to maximum of 16
– More ports equate to more concurrent I/O driven by Spectrum Virtualize
– Support for mapping controller A ports to Fabric A and controller B ports
to Fabric B or cross connecting ports to both fabrics from both
controllers
• IBM® DS4000® does not support a storage system A and storage system B in the
same zone
• Later is preferred to avoid AVT/Trespass occurring if a fabric or all paths to a fabric fail
– Spectrum Virtualize supports Spectrum Virtualize queue depth change
for CX models
• Drives more I/O per port per Mdisk

IBM.
Storage Zones - General Spectrum Virtualize Best Practice Zoning
Concepts
• Create two cluster zones (do not include MM/GM and intra-cluster traffic
ports)
• Never span zones to include more than one Backend storage device!
• Never put Host OS ports, Spectrum Virtualize ports and Backend Storage
ports together in the same zone
– Instead
• Create zones with Host ports and Spectrum Virtualize ports
• Create zones with Backend and Spectrum Virtualize ports
– Never use the same DS8K ports or any native back-end port for
connectivity to Spectrum Virtualize and an attached host
• If Spectrum Virtualize is attached to the DS8K or other native back-end
devices and the DS8K or other back-end device is using native GM (not
Spectrum Virtualize GM) then dedicate appropriate back-end ports
specifically for GM, not to be used for attaching any other device, whether
Host Server, Spectrum Virtualize or other connectivity relationships.

IBM.
Host Zones
Best Practices Concepts
• Too many paths to a Vdisk
– If the recommended number of paths to a Vdisk are exceeded, path failures may not be recovered in the
required amount of time
• Causes excessive I/O waits, resulting in application failures
• Under certain circumstances, it can reduce performance
– Note: 8 paths are supported but 4 are optimum for SDD/SDDDSM/SDDPCM
• Spectrum Virtualize host zones
– There must be a single zone for each host port. This zone must contain the host port, and one port from each
Spectrum Virtualize node that the host will need to access. While there are two ports from each node per SAN
fabric in a usual dual-fabric configuration, make sure that the host only accesses one of them.
• Check with the Host System Administrators to receive output from the SDD or
PCM path query device commands on each host to verify proper pathing
between the Spectrum Virtualize nodes and host systems
– Note: It is a supported configuration to have eight paths to each VDisk, but this design
provides no performance benefit, and it does not improve reliability or availability by any
significant degree

IBM.

IBM Spectrum Virtualize - BP Zoning 101-V9.2

Uploaded by

Copyright:

Available Formats

IBM Spectrum Virtualize - BP Zoning 101-V9.2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IBM Spectrum Virtualize - BP Zoning 101-V9.2

Uploaded by

Copyright:

Available Formats

IBM Spectrum Virtualize Best Practice Zoning 101

• By the end of this session you should know:

• Basic “ground rules and achieving BP rules using:

© Copyright IBM Corporation 2016. Technical University/Symposia materials 1

IBM Spectrum Virtualize Zoning - Best Practices 101

© Copyright IBM Corporation 2016. Technical University/Symposia materials 2

Reason for Best Practice:

DH8 1 Building Block – 16 FC ports per node

Reason for Best Practice:

© Copyright IBM Corporation 2016. Technical University/Symposia materials 5

DH8 1 Building Block – 12 FC ports per node

© Copyright IBM Corporation 2016. Technical University/Symposia materials 6

© Copyright IBM Corporation 2016. Technical University/Symposia materials 7

• The MM/GM port mask (partner_fc_mask)

• Hosts with four (or more) Host Bus Adapters (HBAs)

© Copyright IBM Corporation 2016. Technical University/Symposia materials 9

• The top10 Spectrum Virtulize (SVC) zoning tips:

© Copyright IBM Corporation 2016. Technical University/Symposia materials 10

IBM Spectrum Virtualize Zoning - Best Practices 101

© Copyright IBM Corporation 2016. Technical University/Symposia materials 11

• Use together for full duplex bandwidth

DH8 1 Building Block – 16 FC ports per node

Do not separate port usage, Use the

DH8 1 Building Block – 16 FC ports per node

IBM Spectrum Virtualize Zoning - Best Practices 101

© Copyright IBM Corporation 2016. Technical University/Symposia materials 14

• Should be (1:2) zoning Host

Fabric1 Edge1 Fabric2 Edge1

DH8 1 Building Block – 16 FC ports per node

• No (1:1) zoning Host

Fabric1 Edge1 Fabric2 Edge1

DH8 1 Building Block – 16 FC ports per node

IBM Spectrum Virtualize Zoning - Best Practices 101

© Copyright IBM Corporation 2016. Technical University/Symposia materials 17

• 2 preferred – 1 red & 1 blue Host

Fabric1 Edge1 Fabric2 Edge1

Preferred Paths Non Preferred Paths

DH8 1 Building Block – 16 FC ports per node

Fabric1 Edge1 Fabric2 Edge1

Preferred Paths Non Preferred Paths

DH8 1 Building Block – 16 FC ports per node

• Spectrum Virtualize - DH8 Host

Fabric1 Edge1 Fabric2 Edge1

Preferred Paths Non Preferred Paths

DH8 1 Building Block – 16 FC ports per node

Fabric_A Zone Definitions Fabric_B Zone Definitions

DIR1 SAN Fabric DIR2 SAN Fabric

In this example we show

© Copyright IBM Corporation 2016. Technical University/Symposia materials 21

b03vio210_1 b03vio210_2 b03vio210_3 b03vio210_4

DIR1 SAN Fabric DIR2 SAN Fabric

SVC Host Definitions

© Copyright IBM Corporation 2016. Technical University/Symposia materials 22

• Example of correct/incorrect number of datapaths with Host to Spectrum Virtualize

© Copyright IBM Corporation 2016. Technical University/Symposia materials 23

IBM Spectrum Virtualize Zoning - Best Practices 101

© Copyright IBM Corporation 2016. Technical University/Symposia materials 24

DH8 1 Building Block – 16 FC ports per node

WAN direct connection to the WAN direct connection to the

DH8 1 Building Block – 16 FC ports per node