6500 Backplane

[Adnan Khan]

Hi, 

In 6513 if we have two SUP, one will be active another one will be standby. in 6500 switches backplane is lies on SUP engine not like Nexus. so if we have two SUPs one active other one is standby so only one SUP will provide backplane capacity?

KR,

[Mark Malone]

Hi

if one sup is active you will get the throughput just from that sup if that's what your asking ? the 2 sups don't combine for throughput as one is standby from my understanding of it anyway, that's how ive read it through the docs

http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/prod_white_paper0900aecd80673385.html

Cisco Catalyst 6500 Backplane

The Cisco Catalyst 6500 incorporates two backplanes. From its initial release in 1999, the Cisco Catalyst 6500 chassis has supported a 32-Gbps shared switching bus, a proven architecture for interconnecting line cards within the chassis. The Cisco Catalyst 6500 chassis also includes a second backplane that allows line cards to connect over a high-speed switching path into a crossbar switching fabric. The crossbar switching fabric provides a set of discrete and unique paths for each line card to both transmit data into and receive data from the crossbar switching fabric. The first generation switching fabric was delivered by the switch fabric modules (WS-C6500-SFM and WS-C6500-SFM2), each providing a total switching capacity of 256 Gbps. More recently, with the introduction of the Supervisor Engine 720, the crossbar switch fabric has been integrated into the Supervisor Engine 720 baseboard itself, eliminating the need for a standalone switch fabric module. The capacity of the new integrated crossbar switch fabric on the Supervisor Engine 720 has been increased from 256 Gbps to 720 Gbps. The Supervisor Engine 720-3B and Supervisor Engine 720-3BXL also maintain the same fabric capacity size of 720 Gbps.

Depending on the Cisco Catalyst 6500 chassis, the crossbar switching fabric maps out a series of fabric channels (otherwise known as paths into the crossbar) to each line-card slot in a slightly different layout. Each chassis fabric layout is detailed in Table 1.

Table 1.       Chassis Slot Options

Chassis	Supervisor Engine 32/ Supervisor Engine 720 Slots	Classic Line-Card Slots	Single Fabric Connected Line Cards	Dual Fabric Connected Line Cards
Cisco Catalyst 6503	1 and 2	2 and 3	2 and 3	WS-X67XX line cards not supported
Cisco Catalyst 6503-E	1 and 2	2 and 3	2 and 3	2 and 3
Cisco Catalyst 6504-E	1 and 2	2 thru 4	2 thru 4	2 thru 4
Cisco Catalyst 6506	5 and 6	1 through 6	1 through 6	1 through 6
Cisco Catalyst 6506-E	5 and 6	1 through 6	1 through 6	1 through 6
Cisco Catalyst 6509	5 and 6	1 through 9	1 through 9	1 through 9
Cisco Catalyst 6509-E	5 and 6	1 through 9	1 through 9	1 through 9
Cisco Catalyst 6509-NEB	5 and 6	1 through 9	1 through 9	1 through 9
Cisco Catalyst 6509-NEB-A	5 and 6	1 through 9	1 through 9	1 through 9
Cisco Catalyst 6513	7 and 8	1 through 13	1 through 13	9 through 13

In all but the thirteen slot chassis, each line-card slot has two channels in and out of the switching fabric. The thirteen slot chassis has one fabric channel to each slot in slots 1 through 8 and two fabric channels to each slot in slots 9 through 13. The crossbar switching fabric allows each line card to forward and receive data to every other line card over a unique set of transmission paths.

[Adnan Khan]

Thanks Mark really helpful. One more thing. In Nexus we have 10G interfaces up to 48 port on F2e Line cards. Do we have 10G support on 6500 with all dedicated 10G interfaces or their bandwidth is  shared based on port grouping. 

[Mark Malone]

its shared , take a look at this for 6500 blades last updated start of January last , most are ratio based , I don't see any 48 port blades available though , max 16 , maybe there some in the pipeline but I think the current sups may struggle with a full 48x10gb that's why they may not be available for the 6500

http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/product_data_sheet09186a00801dce34.html

Main Features and Benefits

Table 1 summarizes the primary features and benefits of the Cisco Catalyst 6500 Series 10 Gigabit Ethernet modules.

Table 1.       Cisco Catalyst 6500 Series 10 Gigabit Ethernet Modules Primary Features Comparison

Feature	4-Port 10GbE Fiber Module	8-Port 10GbE Fiber Module	16-Port 10GbE Fiber Module	16-Port 10GbE Copper Module
Ports	4	8	16	16
Optics	XENPAK	X2	X2	●  No optics ●  Copper (RJ-45) connectors
Switch fabric connection	40 Gbps (80 Gbps full duplex)	40 Gbps (80 Gbps full duplex)	40 Gbps (80 Gbps full duplex)	40 Gbps (80 Gbps full duplex)
Oversubscription	1:1	2:1	4:1	4:1
Forwarding engine	●  Default: Centralized forwarding card (CFC) ●  Optional: Distributed forwarding card with DFC3A, DFC3B, DFC3BXL, DFC3C, or DFC3CXL	●  WS-X6708-10G-3C: equipped with DFC3C for distributed forwarding, supporting 256,000 routes ●  WS-X6708-10G-3CXL: equipped with DFC3CXL for distributed forwarding, supporting 1 million routes	●  WS-X6716-10G-3C: equipped with DFC3C for distributed forwarding, supporting 256,000 routes ●  WS-X6716-10G-3CXL: equipped with DFC3CXL for distributed forwarding, supporting 1 million routes	●  WS-X6716-10T-3C: equipped with DFC3C for distributed forwarding, supporting 256,000 routes ●  WS-X6716-10T-3CXL: equipped with DFC3CXL for distributed forwarding, supporting 1 million routes
Queues	●  Receive: 8q8t ●  Transmit: 1p7q8t	●  Receive: 8q4t ●  Transmit: 1p7q4t	Oversubscription mode: ●  Receive: 1p7q2t per port ●  Transmit: 1p7q4t per port group Performance mode: ●  Receive: 8q4t per port ●  Transmit: 1p7q4t per port	Oversubscription mode: ●  Receive: 1p7q2t per port ●  Transmit: 1p7q4t per port group Performance mode: ●  Receive: 8q4t per port ●  Transmit: 1p7q4t per port
Queuing mechanisms	●  Class of service (CoS)–based queue mapping	●  CoS-based queue mapping ●  Differentiated services code point (DSCP)–based queue mapping	●  CoS-based queue mapping ●  DSCP-based queue mapping	●  CoS-based queue mapping ●  DSCP-based queue mapping
Scheduler	●  Deficit Weighted Round Robin (DWRR) ●  Weighted Random Early Detection (WRED)	●  DWRR ●  WRED ●  Shaped Round Robin (SRR) at egress	Oversubscription mode: ●  DWRR ●  WRED Performance mode: ●  DWRR ●  WRED ●  SRR at egress	Oversubscription mode: ●  DWRR ●  WRED Performance mode: ●  DWRR ●  WRED ●  SRR at egress
Port buffers	16 MB per port	200 MB per port	Oversubscription mode: ●  90 MB per port group Performance mode: ●  200 MB per port	Oversubscription mode: ●  90 MB per port group Performance mode: ●  200 MB per port
Hardware-based multicast replication (Layer 2)	●  Ingress and egress ●  Approximately 20 GB per replication engine ●  2 replication engines per module	●  Ingress and egress ●  Approximately 20 GB per replication engine ●  2 replication engines per module	●  Ingress and egress ●  Approximately 20 GB per replication engine ●  2 replication engines per module	●  Ingress and egress ●  Approximately 20 GB per replication engine ●  2 replication engines per module
Hardware-based multicast replication (Layer 3)	●  Ingress and egress ●  Approximately 10 GB per replication engine ●  2 replication engines per module	●  Ingress and egress ●  Approximately 20 GB per replication engine ●  2 replication engines per module	●  Ingress and egress ●  Approximately 20 GB per replication engine ●  2 replication engines per module	●  Ingress and egress ●  Approximately 20 GB per replication engine ●  2 replication engines per module
Jumbo frame support for bridged and routed packets	Up to 9216 bytes	Up to 9216 bytes	Up to 9216 bytes	Up to 9216 bytes
Maximum port density per chassis	34 ports (9-slot chassis)	66 ports (9-slot chassis)	130 ports (9-slot chassis)	130 ports (9-slot chassis)
Maximum port density per VSS	68 ports	132 ports	260 ports	260 ports
Can be used to form virtual switch link	No	Yes	●  Performance mode: Yes (supported in a subsequent software release) ●  Oversubscription mode: No	●  Performance mode: Yes (supported in a subsequent software release) ●  Oversubscription mode: No
Supervisor engines supported	Cisco Catalyst 6500 Series Virtual Switching Supervisor Engine 720 with 10GE uplinks or Supervisor Engine 720 with any policy feature card (PFC; chassis will work in lowest common denominator mode)	Cisco Catalyst 6500 Series Virtual Switching Supervisor Engine 720 with 10GE uplinks or Supervisor Engine 720 with any PFC (chassis will work in lowest common denominator mode)	Cisco Catalyst 6500 Series Virtual Switching Supervisor Engine 720 with 10GE uplinks or Supervisor Engine 720 with any PFC (chassis will work in lowest common denominator mode)	Cisco Catalyst 6500 Series Virtual Switching Supervisor Engine 720 with 10GE uplinks or Supervisor Engine 720 with any PFC (chassis will work in lowest common denominator mode)
Chassis supported	●  Any Cisco Catalyst 6500 E-Series chassis, C6509-NEB-A chassis, non-E-Series chassis with fan tray 2, or Cisco 7600 Series or 7600-S Series chassis (NEBS compliant: operating temperature up to 55°C) ●  Not supported in Cisco Catalyst 6503 non-E Series chassis	●  Any Cisco Catalyst 6500 E-Series chassis, including 6503-E, 6504-E, 6506-E, 6509-E, 6509-V-E, and C6509-NEB-A chassis with dual fan tray, or the Cisco 7604, 7609 with dual fan tray, or 7600-S Series chassis (NEBS compliant: operating temperature up to 55°C), or; ●  Non-E-Series chassis with fan tray 2, including Cisco Catalyst 6506, 6509, 6513, and C6509-NEB-A with single fan tray or the Cisco 7606, 7613, and 7609 chassis with single fan tray (non-NEBS compliant: operating temperature up to 40°C) ●  Not supported in Cisco Catalyst 6503 non-E Series chassis	●  Any Cisco Catalyst 6500 E-Series chassis, including 6503-E, 6504-E, 6506-E, 6509-E, 6509-V-E, and C6509-NEB-A chassis with dual fan tray (NEBS compliant: operating temperature up to 55°C), or; ●  Non-E-Series chassis with fan tray 2, including Cisco Catalyst 6506, 6509, 6513, and C6509-NEB-A with single fan tray (non-NEBS compliant: operating temperature up to 40°C) ●  Not supported in Cisco Catalyst 6503 non-E Series chassis and 7600 Series chassis	●  Any Cisco Catalyst 6500 E-Series chassis, including 6503-E, 6504-E, 6506-E, 6509-E, 6509-V-E, and C6509-NEB-A chassis with dual fan tray (NEBS compliant: operating temperature up to 55°C), or: ●  Non-E-Series chassis with fan tray 2, including Cisco Catalyst 6506, 6509, 6513, and C6509-NEB-A with single fan tray (non-NEBS compliant: operating temperature up to 40°C) ●  Not supported in Cisco Catalyst 6503 non-E Series chassis and 7600 Series chassis
Slot requirements	●  Can occupy any slot in any Cisco Catalyst 6503-E, 6504-E, 6506, 6506-E, 6509, 6509-E, 6509-V-E, or 6509-NEB-A chassis; or Cisco 7604, 7607, 7609, or 7600-S Series chassis ●  Can occupy only slots 9 through 13 in a Cisco Catalyst 6513 or Cisco 7613 chassis	●  Can occupy any slot in any Cisco Catalyst 6503-E, 6504-E, 6506, 6506-E, 6509, 6509-E, 6509-V-E, or 6509-NEB-A chassis; or Cisco 7604, 7606, 7609, or 7600-S Series chassis ●  Can occupy only slots 9 through 13 in a Cisco Catalyst 6513 or Cisco 7613 chassis	●  Can occupy any slot in any Cisco Catalyst 6503-E, 6504-E, 6506, 6506-E, 6509, 6509-E, 6509-V-E, or 6509-NEB-A chassis ●  Can occupy only slots 9 through 13 in a Cisco Catalyst 6513 chassis	●  Can occupy any slot in any Cisco Catalyst 6503-E, 6504-E, 6506, 6506-E, 6509, 6509-E, 6509-V-E, or 6509-NEB-A chassis ●  Can occupy only slots 9 through 13 in a Cisco Catalyst 6513 chassis
Onboard memory	256 MB default, upgradable to 512 MB or 1 GB	1 GB default	1 GB default	1 GB default

[Adnan Khan]

Thanks Mark

[Joseph W. Doherty]

The 6500E chassis is limited to 80Gbps per slot.  I.e., in theory you can support up to 8 10g ports w/o sharing bandwidth.

Some (all?) of the higher density 6500 10g port line cards support an optional "performance mode".  This mode deactivates ports to bring the subscription ratio down, sometimes to 1:1.  When not active, yes, ports on higher density 6500 10g port line cards are in groups that share bandwidth.

BTW, the newer 6807 has a much higher bandwidth design for its line cards slots, but no current supervisor, even the 6T, supports the maximum bandwidth the chassis allows.

[Adnan Khan]

Thanks Joseph

[Joseph W. Doherty]

In all but the thirteen slot chassis, each line-card slot has two channels in and out of the switching fabric. The thirteen slot chassis has one fabric channel to each slot in slots 1 through 8 and two fabric channels to each slot in slots 9 through 13.

Actually the 6513E supports two fabric channels to all slots.  The fabric channel limitation you describe isn't a limitation of the chassis, it's a limitation of the sup720 fabric.  The sup2T supports the 6513E's extra fabric paths, it also supports double the bandwidth, i.e. 80Gbps to all 6513E slots.

[Mark Malone]

Thats good to know thanks I didn't even see anything about the E series for that model in that doc when I was reading it

[Joseph W. Doherty]

When you speak backplane that's usually taken to mean just the chassis connections that "glue" inserted line cards together.  As far as I know, there's no (6500) redundancy for that.  Mark's post better explains.

When a supervisor or supervisors are inserted they support operation of part of the chassis backplane as a system bus.

Later supervisors (i.e. sup 720 and later) provide a switch fabric.  However switch fabric cards (one or two) could be optionally used with the sup2.

With fabric, whether provided by optional fabric cards or supervisiors, you may have one or two, but with two, only one  is actively used.  The other is in standby.