Cake - FQ_codel the next generation
 help / color / mirror / Atom feed
* [Cake] Cobalt testing
@ 2017-09-26  0:15 George Amanakis
  0 siblings, 0 replies; only message in thread
From: George Amanakis @ 2017-09-26  0:15 UTC (permalink / raw)
  To: cake

I am currently testing the cobalt branch with limited rates. The line is 
a residential cable connection (Comcast) advertised as 10Mbps (actual 
speeds are ~12500kbps ingress and ~2600kbps egress). IMO the limited 
rate is more stable after the last commits. I used the apu2.sh and 
dancer.sh scripts (from Dave, http://blog.cerowrt.org/flent/fq_host2/) 
on two different machines simultaneously. BTW, I think dancer.sh has a 
"bug", the last 2 tests should begin with "flent -6" since we want to 
test IPv6.

Config:
qdisc cake 8002: dev ens3 root refcnt 9 bandwidth 2500Kbit diffserv3 
dual-srchost nat wash rtt 100.0ms noatm overhead 18 via-ethernet mpu 64
qdisc cake 8001: dev ens4 root refcnt 9 bandwidth 12200Kbit diffserv3 
dual-dsthost wash ingress rtt 100.0ms noatm overhead 18 via-ethernet mpu 64

Due to the long nature of the results, they are to be found at 
https://pastebin.com/zRvUeCDd.

The only comment I have is that while ingress bandwidth is distributed 
consistently equally between the two hosts (eg 5.81Mbits/s for each one 
in IPv4), egress bandwidth could have been distributed better (eg 
1.06Mbits/s for apu2 vs 1.44Mbit/s for dancer).

George

.

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2017-09-26  0:15 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-09-26  0:15 [Cake] Cobalt testing George Amanakis

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox