This counter tallies the number of received bytes that this PM instance discarded on the way in, as stated in the MIB description. Because this is the total receive-drop counter for the whole interface rather than a per-queue figure, a rising count here points to an ingress-side problem -- a saturated policer, an oversubscribed CIR/EIR rate limit, or a hardware buffer that can't absorb a burst -- happening before traffic ever reaches the egress queues. For example, if this jumps by 119,887 bytes between two consecutive polls while cienaCesPmBasicRxTotalStatsBinBasicRxBytesPerSec stays flat, that's a sign the ingress policer is clipping bursts rather than the link itself losing capacity.