Interpreting morsectrl stats retry table

General
1

I have some questions about the retry table, pulled from the access point.

What is the time scale for average time? Does that represent the time that the packet is waiting before transmission? Is the maximum limit here set by dot11ShortRetryLimit, and if so, can that be changed? Why is there a gap between 8-11 retry count? After 12 retries is the packet dropped?

Retry table
Retry    Count    Avg Time
=====    =====    ========
0        11968062 10637
1        1263021  13734
2        595261   17715
3        219942   22604
4        111364   26996
5        41202    31547
6        20467    37869
7        4149     40117
8        0        0
9        0        0
10       0        0
11       0        0
12       819957   17150
2

Hi @alexb

  • The timescale is in microseconds (us)
  • Avg Time is the average time from when the firmware receives a packet, until it is successfully consumed (i.e. transmitted successfully or failed)
  • MMRC is limited to 7 attempts, which explains the gap between 8-11
    • 12th retry is propagating the failure up to the stack
3

Hello,

Thank you for the response.

I pulled this retry table off another device.

Retry	Count	Avg Time
=====	=====	========
0		14046	14587
1		585		24027
2		295		20901
3		33		47199
4		0		0
5		0		0
6		0		0
7		0		0
8		0		0
9		0		0
10		0		0
11		0		0
12		0		0

How do I reconcile that packets that are retried once, spend more time on average being sent than packets retried twice (24ms vs 20ms)

I’ve also seen retry tables where the average time for x retries is an order of magnitude or two larger than the avg time for x-1 and x+1 retries

4

Hi,

It is likely that powersave is causing this behaviour, where it is asleep for the first retry but awake for the second.