The high watermark level in these queues indicates that the packet processing rate is lower than the incoming rate.
The following example shows output of the dispcnt -p netif -s len -s wmark -d vcgw.com command.
# dispcnt -p netif -s len -s wmark -d vcgw.com Wed Jul 20 09:03:39 2022 netif_queue0_len = 0 0 /s netif_queue0_wmark = 0 0 /s netif_queue0_wmark_1min = 0 0 /s netif_queue0_wmark_1s = 0 0 /s netif_queue0_wmark_5min = 0 0 /s netif_queue1_len = 0 0 /s netif_queue1_wmark = 45 22 /s netif_queue1_wmark_1min = 3 1 /s netif_queue1_wmark_1s = 1 0 /s netif_queue1_wmark_5min = 3 1 /s
In the output, the value in the first column is the current count, and the second column is the rate per second. When you run this command for the first iteration, the first column displays the total (lifetime) count since the counter started, and the second column displays the rate which is calculated by dividing the lifetime total by 2. As the data refreshes every 2 seconds, then the first column displays the count since the last sample, and the second column is the rate per second.
The following example shows output of the dispcnt -p per_core -s len -s wmark -d vcgw.com command.
# dispcnt -p per_core -s len -s wmark -d vcgw.com Wed Jul 20 09:11:05 2022 per_core_queue0_len = 0 0 /s per_core_queue0_wmark = 1476 738 /s per_core_queue0_wmark_1min = 91 45 /s per_core_queue0_wmark_1s = 34 17 /s per_core_queue0_wmark_5min = 346 173 /s per_core_queue1_len = 0 0 /s per_core_queue1_wmark = 1216 608 /s per_core_queue1_wmark_1min = 47 23 /s per_core_queue1_wmark_1s = 27 13 /s per_core_queue1_wmark_5min = 64 32 /s
