Okay so what I am concluding from this discussion is that.
a. the USG series is not capable of tagging 802.1p (L2)
b. the USG series IS capable of tagging with DSCP. (L3)
c. DSCP is not related to 802.1p in the sense that in my twisted mind I could not program lets say half of DSCP or some bits and bytes in DSCP that would magically = 802.1p (in other words its not part of the DSCP tagging bits at all)
d. the netgear GS108T is exactly what I need because it specifically states changing CoS.
Hopefully you can understand my confusion when both the NetGear and USG have a diffserv section in their respective UG and the netgear talks about CoS and the USG talks about ToS and DSCP.
This quote muddied the water for me:
From the Netgear GS108Tv2 Manuel:
"An 802.1p header is inserted if it does not already exist. This is useful for assigning a layer 2 priority level based on a DiffServ forwarding class (i.e., DSCP or IP Precedence value) definition to convey some QoS characteristics to downstream switches which do not routinely look at the DSCP value in the IP header."
Reading the QoS chapter in the Netgear it does have a clear QoS 802.1p chapter but in this basic case its like all the other switches (incl GS2200 series) in that its simply for mapping an internal priority scheme (4 levels) 2:1 to 802.1p standards levels (0-7) so that the switch handles tagged packets accordingly (as the admin intends) and also what level to apply to untagged packets.
Quote: "To map 802.1p priorities to queues" unquote.
But it also states: "Use the DSCP to Queue Mapping page to specify which internal traffic class to map the corresponding DSCP value"
Meaning it doesnt apply any DSCP values and I dont see anywhere it can actually read them, but this switch much like 802.1p allows the admin to allow one to map any DSCP value to one of the four priority schemas (queues) similar to the CoS above.
At the start of the Netgear DIFF Serv chapter it states:
"The QoS feature contains Differentiated Services (DiffServ) support that allows traffic to be classified into streams and given certain QoS treatment in accordance with defined per-hop behaviors." Gobblity gook speak to me.
What one seems to have to do is the following:
1. Class: Create classes and define class criteria.
2. Policy: Create policies, associate classes with policies, and define policy statements.
3. Service: Add a policy to an inbound interface
Packet processing begins by testing the class match criteria for a packet. A policy is applied to a packet when a class match within that policy is found.
In the CLASS menu the following parameter might be germane.
IP Precedence - Matches the packets IP Precedence value to the class criterias when Enter a value in the range of 07. and IP ToS. Matches the packets Type of Service bits in the IP header to the class criterias when selected and a value is entered. In the ToS Bits field, enter a two-digit hexadecimal number to match the bits in a packets ToS field. In the ToS Mask field, specify the bit positions that are used for comparison against the IP ToS field in a packet.
So Class lets you identify the traffic of interest, policy which is next says what your going to do with that traffic.
The following section seems to indicate that this switch can actually change data packet contents (headers).
Configure the policy attributes:.
Drop. Select this option to drop packets for this policy-class.
Mark CoS. Enter the specified Class of Service queue number to mark all packets for the associated traffic stream with the specified class of service value in the priority field of the 802.1p header. If the packet does not already contain this header, one is inserted. The CoS value is an integer from 07.
Mark IP Precedence. Use this attribute to mark all packets for the associated traffic stream with the IP Precedence value you enter in the IP Precedence Value field.
Mark IP DSCP. Use this attribute to mark all packets for the associated traffic stream with IP DSCP value you choose from the menu.
BUT my concern now is that in a following text portion it uses wording to state that all three describes them (CoS, Precedence, & DSCP) in the same terms.
I know it only states in the CoS portion prior "insert a Cos Value" and I also know that this L2 switch cannot enter L2 DSCP parameters into the data. So I am confused...........
Violate Action. Determines what happens to packets that are considered non-conforming (above the police rate). Select one of the following actions:
Send. (default) These packets are presented unmodified by DiffServ to the system forwarding element.
Drop. (default) These packets are immediately dropped.
Mark CoS. These packets are marked by DiffServ with the specified CoS value before being presented to the system forwarding element. This selection requires that the Mark CoS value field be set.
Mark IP Precedence. These packets are marked by DiffServ with the specified IP Precedence value before being presented to the system forwarding element. This selection requires that the Mark IP Precedence value field be set.
Mark IP DSCP. These packets are marked by DiffServ with the specified DSCP value before being presented to the system forwarding element. This selection requires that the DSCP value field be set.
So is CoS being marked mean the same as IP Precendence and IP DSCP being marked or is the COS different in that it can be inserted???
I am assuming that switch can only change parameters at Level 2, and why only CoS is noted as having insertion capability, which is what I need and thus my conclusion on the Netgear as a viable platform for what I am researching is still valid.
To mozerd, the switch would be configured as follows:
port 1 WAN incomming from the the ONT - VLAN 34, VLAN35
Port 2-X connetivity to STBs - VLAN 34 Diff Serv CoS 4
Port 8 To WAN port on USG - VLAN 35--
Ain't nuthin but the blues! "Albert Collins".
Leave your troubles at the door! "Pepe Peregil" De Sevilla. Just Don't Wifi without WPA, "Yul Brenner"