User Rating: 0 / 5

Star InactiveStar InactiveStar InactiveStar InactiveStar Inactive
 

Introduction

Configuration

Testing

Summary

Introduction

The constant development of wireless networks forces network administrators to unceasingly upgrade their wired segments supporting operation of Wi-Fi. For now, during the usage of high speed access points, wired interfaces of access level switches begin to become a bottleneck, restricting the data transfer rates in a wireless segment. It seems incredible, but even a gigabit interface is occasionally not enough. However, the transition to the use of 10GE networks won't be considered justified for a long time as so far the speeds provided by Wi-Fi access points don't plan to cross this limit.

Foreseeing the aggravation of the situation in a wired segment in connection with the appearance of the devices with 802.11AC Wave 2 support, Cisco Systems, Inc. together with a number of other vendors founded NBASE-T alliance in 2014. The purpose of this organization became the development of Ethernet standards, that allow realizing data transfer at 2.5 and 5 Gbps, using the existing cabling infrastructure of 5e and 6 categories.

What does the appearance of the devices with "the second wave" of 802.11AC standard support threaten network administrators with? First, there will be an increase in the maximum theoretical data transfer rates to 6.8 Gbps. Of course, it is only a theoretical maximum (actual speeds will be traditionally half as high), moreover, achievable only when using the most efficient clients located in close proximity to the access point. The second improvement in 802.11AC Wave 2 standard is the support of MU-MIMO technology. Usage of MU-MIMO will allow distributing more effectively available bandwidth between several wireless clients working simultaneously. So, for example, the access point with 4x4 antenna configuration will be able to service two 2x2 clients at the same time, rather than sequentially, as it was before.

Both improvements available in 802.11AC Wave 2 will lead to a much higher utilization of the wired network segments. NBASE-T standards, enabling to prepare for implementation of the wireless equipment with support of IEEE 802.11AC Wave 2 with the minimum expenses, were developed to eliminate bottlenecks in modern L2-segments. Besides, existence of Power over Ethernet technology support, enabling to realize a remote powering of access points, surveillance cameras and other network equipment, should not be left unnoticed in NBASE-T.

Today in our test laboratory there is Cisco Catalyst WS-C3560CX-8XPD-S switch (IOS version 15.2 (6) E), with two multigigabit interfaces. The specified interfaces support the following transmission speeds: 100 Mbps, 1 Gbps, 2.5 Gbps, 5 Gbps and 10 Gbps. Of course, two multigigabit ports are not the only switch interfaces. 3560CX-8XPD model is equipped with six Gigabit Ethernet ports and also two connectors for SFP+ modules. All eight copper interfaces support PoE+ transmission, the switch's energy budget is 240W.

Configuration

Cisco 3560CX-8XPD Switch has four 10GE interfaces: Te1/0/1, Te1/0/2, Te1/0/7 and Te1/0/8, the two latter specifically have NBASE-T support.

fox_3560CX-8XPD#sho int status
Port Name Status Vlan Duplex Speed Type
Gi1/0/1 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/2 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/3 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/4 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/5 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/6 notconnect 1 auto auto 10/100/1000BaseTX
Te1/0/7 connected 1 a-full 100 100/1G/2.5G/5G/10GBaseT
Te1/0/8 connected 1 a-full 100 100/1G/2.5G/5G/10GBaseT
Te1/0/1 notconnect 1 full 10G Not Present
Te1/0/2 notconnect 1 full 10G Not Present

Multigigabit interfaces don't require any additional configuration – even speed can be determined automatically.

fox_3560CX-8XPD(config)#int te1/0/7
fox_3560CX-8XPD(config-if)#?
Interface configuration commands:
 aaa Authentication, Authorization and Accounting.
 access-session Access Session specific Interface Configuration Commands
 arp Set arp type (arpa, probe, snap) or timeout or log options
 auto Configure Automation
 auto Configure Automation
 bandwidth Set bandwidth informational parameter
 bfd BFD interface configuration commands
 bgp-policy Apply policy propagated by bgp community string
 carrier-delay Specify delay for interface transitions
 cdp CDP interface subcommands
 channel-group Etherchannel/port bundling configuration
 channel-protocol Select the channel protocol (LACP, PAgP)
 crypto Encryption/Decryption commands
 cts Configure Cisco Trusted Security
 dampening Enable event dampening
 datalink Interface Datalink commands
 default Set a command to its defaults
 delay Specify interface throughput delay
 description Interface specific description
 downshift link downshift feature
 exit Exit from interface configuration mode
 flow-sampler Attach flow sampler to the interface
 flowcontrol Configure flow operation.
 help Description of the interactive help system
 history Interface history histograms - 60 second, 60 minute and 72 hour
 hold-queue Set hold queue depth
 ip Interface Internet Protocol config commands
 ipv6 IPv6 interface subcommands
 isis IS-IS commands
 iso-igrp ISO-IGRP interface subcommands
 keepalive Enable keepalive
 l2protocol-tunnel Tunnel Layer2 protocols
 lacp LACP interface subcommands
 link Interface link related commands
 lldp LLDP interface subcommands
 load-interval Specify interval for load calculation for an interface
 location Interface location information
 logging Configure logging for interface
 mac MAC interface commands
 macro Command macro
 macsec Enable macsec on the interface
 mdix Set Media Dependent Interface with Crossover
 media-proxy Enable media proxy services
 metadata Metadata Application
 mka MACsec Key Agreement (MKA) interface configuration
 mls mls interface commands
 mvr MVR per port configuration
 neighbor interface neighbor configuration mode commands
 network-policy Network Policy
 nmsp NMSP interface configuration
 no Negate a command or set its defaults
 onep Configure onep settings
 ospfv3 OSPFv3 interface commands
 pagp PAgP interface subcommands
 power Power configuration
 priority-queue Priority Queue
 queue-set Choose a queue set for this queue
 rep Resilient Ethernet Protocol characteristics
 rmon Configure Remote Monitoring on an interface
 routing Per-interface routing configuration
 service-policy Configure CPL Service Policy
 shutdown Shutdown the selected interface
 small-frame Set rate limit parameters for small frame
 snmp Modify SNMP interface parameters
 source Get config from another source
 spanning-tree Spanning Tree Subsystem
 speed Configure speed operation.
 srr-queue Configure shaped round-robin transmit queues
 storm-control storm configuration
 subscriber Subscriber inactivity timeout value.
 switchport Set switching mode characteristics
 timeout Define timeout values for this interface
 topology Configure routing topology on the interface
 transmit-interface Assign a transmit interface to a receive-only interface
 tx-ring-limit Configure PA level transmit ring limit
 udld Configure UDLD enabled or disabled and ignore global UDLD setting
 vtp Enable VTP on this interface
fox_3560CX-8XPD(config-if)#speed ?
 100 Force 100 Mbps operation
 1000 Force 1000 Mbps operation
 10000 Force 10000 Mbps operation
 2500 Force 2500 Mbps operation
 5000 Force 5000 Mbps operation
 auto Enable AUTO speed configuration

There is no duplex setting for NBASE-T ports.

fox_3560CX-8XPD(config)#int gi1/0/1
fox_3560CX-8XPD(config-if)#du ?
 auto Enable AUTO duplex configuration
 full Force full duplex operation
 half Force half-duplex operation
fox_3560CX-8XPD(config-if)#int te1/0/7
fox_3560CX-8XPD(config-if)#du ?
% Unrecognized command

As for the Auto MDI/MDIX function operation, in this switch determination of the cable used is made regardless of the speed at which the interface functions.

When using auto negotiation the administrator can explicitly indicate what speeds are acceptable for negotiation. It is fair to say that similar setting is available for gigabit interfaces as well.

fox_3560CX-8XPD(config-if)#spe au ?
 100 Include 100 Mbps in auto-negotiation advertisement
 1000 Include 1000 Mbps in auto-negotiation advertisement
 10000 Include 10000 Mbps in auto-negotiation advertisement
 2500 Include 2500 Mbps in auto-negotiation advertisement
 5000 Include 5000 Mbps in auto-negotiation advertisement

We connected Te1/0/7 and Te1/0/8 interfaces using the patch cord. The output of some show commands is presented below.

fox_3560CX-8XPD#sho run int te1/0/7
Building configuration...
Current configuration : 59 bytes
!
interface TenGigabitEthernet1/0/7
 load-interval 30
end
fox_3560CX-8XPD#sho run int te1/0/8
Building configuration...
Current configuration : 71 bytes
!
interface TenGigabitEthernet1/0/8
 load-interval 30
 speed 5000
end
fox_3560CX-8XPD#sho int status
Port Name Status Vlan Duplex Speed Type
Gi1/0/1 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/2 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/3 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/4 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/5 notconnect 1 auto auto 10/100/1000BaseTX
Gi1/0/6 notconnect 1 auto auto 10/100/1000BaseTX
Te1/0/7 connected 1 a-full a-5000 100/1G/2.5G/5G/10GBaseT
Te1/0/8 connected 1 a-full 5000 100/1G/2.5G/5G/10GBaseT
Te1/0/1 notconnect 1 full 10G Not Present
Te1/0/2 notconnect 1 full 10G Not Present
fox_3560CX-8XPD#sho int te1/0/7
TenGigabitEthernet1/0/7 is up, line protocol is up (connected)
 Hardware is Ten Gigabit Ethernet, address is 9c57.adb0.3487 (bia 9c57.adb0.3487)
 MTU 1500 bytes, BW 5000000 Kbit/sec, DLY 10 usec,
 reliability 255/255, txload 1/255, rxload 1/255
 Encapsulation ARPA, loopback not set
 Keepalive set (10 sec)
 Full-duplex, 5000Mb/s, media type is 100/1G/2.5G/5G/10GBaseT
 input flow-control is off, output flow-control is unsupported
 ARP type: ARPA, ARP Timeout 04:00:00
 Last input 00:00:01, output 00:00:01, output hang never
 Last clearing of "show interface" counters never
 Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
 Queueing strategy: fifo
 Output queue: 0/40 (size/max)
 30 second input rate 0 bits/sec, 0 packets/sec
 30 second output rate 1000 bits/sec, 1 packets/sec
 538 packets input, 102715 bytes, 0 no buffer
 Received 325 broadcasts (325 multicasts)
 0 runts, 0 giants, 0 throttles
 0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
 0 watchdog, 325 multicast, 0 pause input
 0 input packets with dribble condition detected
 1622 packets output, 172091 bytes, 0 underruns
 0 output errors, 0 collisions, 1 interface resets
 0 unknown protocol drops
 0 babbles, 0 late collision, 0 deferred
 0 lost carrier, 0 no carrier, 0 pause output
 0 output buffer failures, 0 output buffers swapped out
fox_3560CX-8XPD#sho int te1/0/8
TenGigabitEthernet1/0/8 is up, line protocol is up (connected)
 Hardware is Ten Gigabit Ethernet, address is 9c57.adb0.3488 (bia 9c57.adb0.3488)
 MTU 1500 bytes, BW 5000000 Kbit/sec, DLY 10 usec,
 reliability 255/255, txload 1/255, rxload 1/255
 Encapsulation ARPA, loopback not set
 Keepalive set (10 sec)
 Full-duplex, 5000Mb/s, media type is 100/1G/2.5G/5G/10GBaseT
 input flow-control is off, output flow-control is unsupported
 ARP type: ARPA, ARP Timeout 04:00:00
 Last input 00:00:01, output 00:00:09, output hang never
 Last clearing of "show interface" counters never
 Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
 Queueing strategy: fifo
 Output queue: 0/40 (size/max)
 30 second input rate 0 bits/sec, 0 packets/sec
 30 second output rate 0 bits/sec, 0 packets/sec
 1626 packets input, 172811 bytes, 0 no buffer
 Received 1413 broadcasts (1397 multicasts)
 0 runts, 0 giants, 0 throttles
 4 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
 0 watchdog, 1397 multicast, 0 pause input
 0 input packets with dribble condition detected
 561 packets output, 104187 bytes, 0 underruns
 0 output errors, 0 collisions, 1 interface resets
 0 unknown protocol drops
 0 babbles, 0 late collision, 0 deferred
 0 lost carrier, 0 no carrier, 0 pause output
 0 output buffer failures, 0 output buffers swapped out
fox_3560CX-8XPD#

Before we proceed directly to the performance testing, we decided to connect the access point with PoE support to Te1/0/7 interface and provide our readers with some diagnostic information.

fox_3560CX-8XPD#sho power inline tenGigabitEthernet 1/0/7
Interface Admin Oper Power Device Class Max
 (Watts)
--------- ------ ---------- ------- ------------------- ----- ----
Te1/0/7 auto on 15.4 Ieee PD 4 30.0
Interface AdminPowerMax AdminConsumption
 (Watts) (Watts)
---------- --------------- --------------------
Te1/0/7 30.0 15.4
fox_3560CX-8XPD#sho lldp ne de
fox_3560CX-8XPD#sho lldp ne detail
------------------------------------------------
Local Intf: Te1/0/7
Chassis id: b8ec.a3ac.5c19
Port id: 1
Port Description: UPLINK
System Name: WAC6103D-I
System Description:
Linux
Time remaining: 118 seconds
System Capabilities: B,W,R
Enabled Capabilities: B,W,R
Management Addresses:
 IP: 192.168.1.2
Auto Negotiation - not supported
Physical media capabilities - not advertised
Media Attachment Unit type - not advertised
Vlan ID: - not advertised
Total entries displayed: 1
fox_3560CX-8XPD(config)#int te1/0/7
fox_3560CX-8XPD(config-if)#po
fox_3560CX-8XPD(config-if)#power ?
 inline Inline power configuration
fox_3560CX-8XPD(config-if)#power i
fox_3560CX-8XPD(config-if)#power inline ?
 auto Automatically detect and power inline devices
 consumption Configure the inline device consumption
 never Never apply inline power
 police Police the power drawn on the port
 port Configure Port Power Level
 static High priority inline power interface

At that the chapter, related to the configuration of multigigabit interfaces, came to an end.

Testing

In this section we want to provide our readers with the results of the switch’s performance tests not only when using multigigabit interfaces, but also using "standard" ports. As a traffic generator IXIA hardware and software system was used. We decided to begin with performance measurement (a bandwidth, a latency and a jitter) of 3560CX-8XPD model executing switching with the use of Gigabit Ethernet interfaces. Measurements were made for frames of different length. During these tests we didn't detect packet losses (excluding IPv6 routing test), therefore we decided not to include this diagram in the article.

Since Cisco Catalyst 3560CX-8XPD model has the feasibility to execute not only switching of Ethernet frames, but also IP packets routing, we decided not to leave this feature without attention.

It can be seen from the diagrams above that the device performance in case of L2 functions execution has much in common with L3 functions execution.

Performance measurement of the switch in case of connection to its 10GE optical ports also in L2 and L3 modes became the next stage.

The switch under test can execute not only routing of IPv4 traffic, but also IPv6. Naturally, we didn't disregard such an opportunity.

It should be noted that when routing 72 bytes packets 0.029% packet loss was observed. The value of losses is small, however we have considered it necessary to mention that.

We approached, perhaps, the most interesting part of this section – switch’s performance measurement using NBASE-T ports. The traffic generator was connected to optical interfaces of the switch (10GE). Multigigabit interfaces were connected to each other with a 0.5 meters patch cord. On the diagrams below a bandwidth, a latency and a jitter for all speeds supported by the interfaces are presented. When plotting the graphs of a latency and a jitter dependence on the packets size we, naturally, considered that frames pass the switch twice in this diagram.

In conclusion of this chapter we would like to provide our readers with the same dependences, but without the use of the switch, that is in a case when traffic generator ports are connected directly to each other.

That's where we draw the testing chapter to a close and move on to summing it all up.

Summary

In this article we considered copper multigigabit interfaces through the example of Cisco Catalyst 3560CX-8XPD compact switch, which has two ports with support of NBASE-T. The use of NBASE-T interfaces will allow updating the existing wired network segments with the minimum expenses and to be prepared for the deployment of the next generation IEEE 802.11AC Wave 2 wireless networks. NBASE-T usage allows increasing significantly the performance of the wired segments without a cable infrastructure replacement. Support of "standard" GE and 10GE speeds by the multigigabit interfaces will allow carrying out changeover of the equipment as smoothly and consistently as possible.

It is also worth noting that the introduction of the new standards with the "intermediate" speeds happened not only in a world of copper network interfaces with RJ45 (8P8C) connectors, but also in the segment of optical networks. An example is Cisco Nexus 3232C switch with high port density, carrying 32 fixed 100GE QSFP28 format interfaces, which allows providing support of the following speeds for optical interfaces: 10G/25G/40G/50G/100G. However that's quite a different story.

Add comment


Security code
Refresh

Found a typo? Please select it and press Ctrl + Enter.