Getting to know the different QOS features of the Linksys Managed Switches

The Linksys Managed Switches, LGS528, LGS528P, LGS552, and LGS552P are designed to customers, who need enhanced Security and advanced QoS. 
 
The article below will show you the different QOS features of the Linksys Managed Switches.
 
Queue Scheduling
 
The Linksys Managed Switches supports four (4) queues for each interface.  Queue 4 is the highest priority queue while Queue 1 is the lowest priority queue.
 
There are two (2) ways of determining how traffic in queues are handled, Strict Priority and Weighted Round Robin (WRR).
  • Strict Priority — The traffic leaving from the highest priority queue is transmitted first.  Traffic from the lower queues is processed only after the highest queue has been transmitted.   This provides the highest level of priority of traffic to the highest numbered queue.
     
  • Weighted Round Robin (WRR) — In this mode, the higher the weight the more frames are sent.  For example, if there are a maximum of four (4) queues possible in WRR using the default weights and all queues are overloaded, it will result to the following bandwidth allocation:
     
  • Queue 1 receives 1/15 of the bandwidth
  • Queue 2 receives 2/15 of the bandwidth
  • Queue 3 receives 4/15 of the bandwidth
  • Queue 4 receives 8/15 of the bandwidth
The type of WRR algorithm used in the device is not the standard Deficit WRR (DWRR), but rather Shaped Deficit WRR (SDWRR).  When the queuing mode is WRR, queues are serviced until their quota has been used up and then another queue is serviced.

The queuing modes can be selected in the Queue Scheduling page.  When the queuing mode is set by Strict Priority, the priority sets the order in which queues are serviced, starting with the highest priority queue and going to the next lower queue when each queue is completed.
 
It is also possible to assign some of the lower queues to WRR, while keeping some of the higher queues in strict priority.  In this case, traffic for the Strict Priority queues is always sent before traffic from the WRR queues.  Only after the Strict Priority queues have been emptied is traffic from the WRR queues forwarded.   (The relative portion from each WRR queue depends on its weight).

Scheduling Method
  • Strict Priority - Traffic scheduling for the selected queue and all higher queues is based strictly on the queue priority.
     
  • Weighted Round Robin Weight - If WRR is selected, enter the WRR weight assigned to the queue.
     
  • %WRR Bandwidth - Displays the amount of bandwidth assigned to the queue.  These values represent the percent of the WRR weight.
Bandwidth Control

The Bandwidth Control page enables users to define two (2) values, Ingress Rate Limit and Egress Shaping Rate.  The Egress Shaping Rate determines how much traffic the system can receive and send.  The Ingress Rate Limit is the number of bits per second that can be received from the ingress interface.  Excess bandwidth above this limit is discarded.
  • Ingress Rate Control
     
  • Ingress Rate Limit - Enter the maximum amount of bandwidth allowed on the interface.  The two (2) Ingress Rate Limit fields do not appear when the interface type is LAG.
     
  • Ingress Committed Burst Size - Enter the maximum burst size of data for the ingress interface in bytes of data.  This amount can be sent even if it temporarily increases the bandwidth beyond the allowed limit.  This field is only available if the interface type is Port.
     
  • Egress Shaping Control
     
  • Egress Committed Information Rate – This sets the average maximum amount of data allowed to be sent on the egress interface, measured in bits per second.   Enter the maximum bandwidth for the egress interface.
     
  • Egress Committed Burst Size – This is the burst of data that is allowed to be sent, even though it is above the committed information rate.  This is defined in the number of bytes of data.  Enter the maximum burst size of data for the egress interface in bytes of data.  This amount can be sent even if it temporarily increases the bandwidth beyond the allowed limit.
Basic QoS Mode

In QoS Basic Mode, a specific domain in the network can be defined as trusted.  Within that domain, packets are marked with 802.1p priority and/or DSCP to signal the type of service they require.  Nodes within the domain use these fields to assign the packet to a specific output queue.  The initial packet classification and marking of these fields is done in the ingress of the trusted domain.

The Linksys Managed Switches supports CoS/802.1p Trusted Mode and DSCP Trusted Mode.  CoS/802.1p Trusted Mode uses the 802.1p priority in the VLAN tag.  DSCP Trusted Mode uses the DSCP value in the IP header. 

An exception to this if that, there is any port should not trust the incoming CoS mark, disable the QoS state on that port using the Feature Configuration page.  Enable or disable the global selected trusted mode at the ports by using the Basic QoS page. 

NOTE:  If a port is disabled without the Trusted Mode, all its ingress packets are forwarded in best effort.  It is recommended that you disable the Trusted Mode at the ports where the CoS/802.1p and/or DSCP values in the incoming packets that are not trustworthy.  Otherwise, it might negatively affect the performance of your network.

Was dit artikel nuttig?

Nog meer vragen?
Opnieuw zoeken