Provision

DSCP Mapping capabilities are provided to identify by DSCP and mark 802.1p. Class of Service (CoS) behavior is governed by the Connection/Service Profile CoS and 802.1p marking. DSCP is not used by itself for service differentiation, rather as mapped and marked 802.1p values. The ability to classify and mark DSCP itself is not provided.

Use the following procedure to enable DSCP Mapping.

1. Configure the map itself, (i.e., what 802.1p values are to be marked by which DiffServ Code Point). There are 64 Code Points. Therefore, there are 64 entries in the DSCP Mapping table; one for each code point. There is often confusion about the Code Point values in this table. Our table uses DSCP values 0-63. In some literature, the Code Points are defined per the entire IP Type of Service (ToS) field, which is 8 bits and the least significant 2 bits are 0. If, using this vernacular, the Code Points are 0, 4, 8, 12, 16, etc., equates to the values in our table of 0, 1, 2, 3, 4, etc., (refer to Create a DSCP Mapping Profile).
2. Configure a Service or Connection Profile to use DSCP Mapping as the 802.1pBit treatment method, (a.k.a Marking Mode). Refer to Ethernet Port Profiles or Managing Connection Profiles. 
3. Then the DSCP profile must be assigned to the ONT. Since DSCP profiles are on an ONT-by-ONT basis, it is best to simply edit the standard Default_DSCP_Map so that it applies to all ONTs, without having to edit each ONT individually to assign the DSCP Mapping Profile to the ONT (refer to the ONT Converting and Activating "Advanced" tab).

4. Uplink DSCP Profile Configuration.

5. When DSCP to pbit marking is required on the uplinks, this can be configured via the LAG Configuration menu.

6. Selecting the drop down for DSCP profile will enable the DSCP to pbit marking on that LAG.  If the DSCP profile is not selected, then packets are passed through the uplink downstream without modifying the pbits.  It is rare that this is needed as typically the L2 marking coming from the network is correct.