Click the exhibit button below. A service provider has applied the SAP-ingress policy configuration below on his customer's SAP. The service provider notices that all of the customer's traffic is…

Question

Exhibit

Options

• ASet the CIR of queue 3 to a non-zero value corresponding to the customer's expected bandwidth
• BSet queue 3 to use priority mode.
• CSet the traffic classification of packets to in-profile.
• DSet the high-priority-only value to 0.
• EMap AF traffic to queue 1

• A
  8% (2)
• B
  12% (3)
• C
  80% (20)

Explanation

In Nokia SR OS, when a SAP-ingress queue has a high-priority-only value set above 0, it reserves queue buffer exclusively for in-profile (high-priority) packets and drops all out-of-profile traffic - which explains why all customer traffic is being dropped. Setting traffic classification to in-profile (C) fixes this because in-profile packets are never restricted by the high-priority-only threshold. Setting high-priority-only to 0 (D) removes the restriction entirely, allowing both in- and out-of-profile packets to use the queue. Mapping AF traffic to queue 1 (E) works because queue 1 presumably lacks the problematic high-priority-only restriction, so traffic no longer hits the broken queue 3.

Why A and B are wrong: A CIR of 0 (A) does not cause drops - it simply means no bandwidth is guaranteed, but traffic can still flow up to the PIR; the high-priority-only setting is the actual culprit. Priority mode (B) changes scheduling behavior between queues but does not affect packet admission or the high-priority-only drop mechanism.

Memory tip: Think of high-priority-only as a VIP-only door policy - if the bouncer only lets in VIPs (in-profile packets), everyone else gets dropped. You can fix it by making everyone a VIP (C), firing the bouncer (D), or using a different door with no policy (E). CIR and scheduling (A, B) are about the queue inside, not who gets past the door.

No community discussion yet for this question.