KSZ9563RNXI datasheet(40/226 Pages) MICROCHIP

40 / 226 page

KSZ9563R

DS00002419D-page 40

 2017-2018 Microchip Technology Inc.

If any egress queue receives more traffic than the specified egress rate throughput, packets may be accumulated in the

output queue and packet memory. After the memory of the queue or the port is used up, packet dropping or flow control

will be triggered. As a result of congestion, the actual egress rate may be dominated by flow control/dropping at the

ingress end, and may be therefore slightly less than the specified egress rate.

To reduce congestion, it is a good practice to ensure that the egress bandwidth exceeds the ingress bandwidth.

4.4.14

EGRESS TRAFFIC SHAPING

The Credit-Based Shaper is defined in IEEE 802.1Qav for Audio Video Bridging (AVB). It attempts to minimize the jitter

of traffic in the egress queue to which it applies, but the transmission of large packets from the “best effort” queue(s) can

negatively impact its performance. When the shaper is used, the corresponding egress ports are normally configured

for either two or four queues. One or more queues may be used for time-sensitive (i.e. scheduled) traffic, and the remain-

ing queue(s) are used for lower priority unscheduled (best effort) traffic. Shaping is individually configurable for each

port and each queue. The Credit-Based Shaper may be applied to either one or two queues on the same port.

Section 4.4.4, "Quality-of-Service (QoS) Priority Support" discusses how ingress packets are mapped to egress queues

based on priority. The most common method is by the PCP field of the VLAN tag.

4.4.14.1

IEEE 802.1Qav Credit-Based Traffic Shaper

The traffic shaper is used to meter high priority (AVB) egress traffic as determined by the reserved bandwidth of the SR

class traffic. A separate traffic shaper is available at the egress of each priority queue at each port. If a port is configured

for four queues, then the two highest priority queues may be used for SR traffic classes. If a port is configured for two

queues, then the highest priority queue (1) may be used for the SR traffic class. The lower priority queue(s) are used

for “best effort” class traffic and are configured for strict priority scheduling, with the traffic shaper disabled for these

queues.

The traffic shaper is more sophisticated than the traditional “leaky bucket” rate limiting feature described earlier. A cer-

tain percentage of a port's bandwidth may be reserved for an AVB stream reservation (SR) traffic class. The traffic

shaper may be configured for the reserved bandwidth, allowing the queue to egress packets at a rate up to but not

exceeding this bandwidth. The queue may accumulate “credits” if an egress packet is delayed by a packet egressing

from another queue. Accumulated credits may be used to permit catching up, in order to maintain an average rate.

If a traffic class supported by the credit-based shaper uses less than the bandwidth allocated to it, then the unused band-

width can be used by other traffic classes, in accordance with the relative priorities of the traffic classes and the trans-

mission queuing algorithms associated with them.

4.4.15

INGRESS MAC ADDRESS FILTERING FUNCTION

When a packet is received, the destination MAC address is looked up in both the static and dynamic MAC address

tables. If the address is not found in either of these tables, then the destination MAC address is “unknown”. By default,

an unknown packet is forwarded to all ports except the port at which it was received. An optional feature makes it pos-

sible to specify the port or ports to which to forward unknown packets. It is also possible to specify no ports, meaning

that unknown packets will be discarded. This feature is implemented separately for unknown unicast, unknown multicast

and unknown VID packets.

4.4.16

802.1X PORT-BASED ACCESS CONTROL

IEEE 802.1X is a Port-based authentication protocol. EAPOL is the protocol normally used by the authentication pro-

cess as uncontrolled Port. By receiving and extracting special EAPOL frames, the host processor can control whether

the ingress and egress ports should forward packets or not. If a user port wants service from another port (authentica-

tor), it must get approved by the authenticator. The device detects EAPOL frames by checking the destination address

of the frame. The destination addresses should be either a multicast address as defined in IEEE 802.1x (01-80-C2-00-

00-03) or an address used in the programmable reserved multicast address domain with offset -00-03. Once EAPOL

frames are detected, the frames are forwarded to the host port so it can send the frames to the authenticator server.

Eventually, the CPU determines whether the requester is qualified or not based on its source MAC address, and frames

are either accepted or dropped.

When the device is configured as an authenticator, the ports of the switch must then be configured for authorization. In

an authenticator-initiated port authorization, a client is powered up or plugs into the port, and the authenticator port

sends an Extensible Authentication Protocol (EAP) PDU to the supplicant requesting the identification of the supplicant.

At this point in the process, the port on the switch is connected from a physical standpoint; however, the 802.1X process

has not authorized the port and no frames are passed from the port on the supplicant into the switching fabric. If the

supplicant attached to the switch (KSZ9563R) did not understand the EAP PDU that it was receiving from the switch, it

KSZ9563RNXI 数据表(PDF) 40 Page - Microchip Technology

KSZ9563RNXI 数据表(HTML) 40 Page - Microchip Technology

Html Pages

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 ...More

数据表下载