High Scale Ethernet MDA Capabilities
Quality of Service Guide 779
• The MBS value is multiplied by the starting-depth percentage to derive the number
of bytes at which the slope will start. This value is then converted to buffers by
dividing the resulting byte value by the buffer size (168 bytes) and rounding to the
nearest integer value. This is the HSMDA slope start-buffer parameter (14 bit value).
• The same process is performed for the maximum-depth percentage to derive the
HSMDA slope end-buffer parameter (14 bit value).
• The system calculates the Inverse-slope value using the following formula:
→ Inverse-slope = (end-buffer — start-buffer) / maximum-discard-probability *
1.27.
• The Inverse-Slope value is converted to an 8 bit fixed point notation where:
→ Bits 7 through 4 represent the value above zero and bits 3 through 0 represent the
fraction below zero.
• If the inverse slope is less than 0.0625, the system uses a value of 0.0625 (0000.0001)
as the 8 bit value.
• If the inverse slope is greater than 15.9375, the system uses 15.9375 (1111.1111) as
the 8 bit value.
• Table 61 can be used to derive the 8-bit value for inverse slopes within the range
0.0625 and 15.9375.
Table 61: HSMDA Inverse Slope Fixed Point Binary Values
Inverse Slope Fixed Point Binary Format
(23 22 21 20. 2-1 2-2 2-3 2-4)
Whole Number Decimal Value Corresponding Binary Value
(23 22 21 20)
00000
10001
20010
30011
40100
50101
60110
70111
81000
91001
To Next Page
Loading...
