* Re: [Bloat] Bismark router + ecn + red tested
2011-05-11 8:49 ` Pedro Tumusok
@ 2011-05-11 13:35 ` Dave Taht
2011-05-11 18:42 ` Jesper Dangaard Brouer
1 sibling, 0 replies; 4+ messages in thread
From: Dave Taht @ 2011-05-11 13:35 UTC (permalink / raw)
To: Pedro Tumusok; +Cc: bloat
[-- Attachment #1: Type: text/plain, Size: 5130 bytes --]
I was using a lightly modified version of nbd's qos-scripts package, which
use red
"lightly modified" meant that I had added options to do SFB and the like,
the result I published earlier was just watching "red" work.
It uses a lot of different qdisc types and it actually turned out to be hard
to get red doing its thing. I am working on something simpler.
On Wed, May 11, 2011 at 2:49 AM, Pedro Tumusok <pedro.tumusok@gmail.com>wrote:
> Hi,
>
> I got a Linux based router, ZyXEL P2612 (ADSL2+) and also the P2812
> (VDSL2) both got an ethernet wan port also.
> I can mess around with qdisc on it, since the web ui QoS setup just
> creates qdisc settings to correspond.
>
> Can I have a copy of the red settings commands you used?
>
> My router basically creates a HTB with 3 SFQ for the QoS and when I
> poked at it, it looked like it would let me change the SFQ to RED.
> But at that time, I did not have the time to really poke around with it.
>
> # tc -s -d qdisc show
> qdisc htb 1: dev eth0 r2q 500 default 10 direct_packets_stat 4 ver 3.17
> Sent 7150514979 bytes 10846403 pkt (dropped 0, overlimits 0 requeues 0)
> rate 154592bit 14pps droprate 0bit backlog 0b 0p requeues 0
> qdisc sfq 10: dev eth0 parent 1:10 limit 128p quantum 1514b flows
> 128/1024 perturb 10sec
> Sent 7150491288 bytes 10846212 pkt (dropped 0, overlimits 0 requeues 0)
> rate 154592bit 14pps droprate 0bit backlog 0b 0p requeues 0
> qdisc sfq 110: dev eth0 parent 1:110 limit 128p quantum 1514b flows
> 128/1024 perturb 10sec
> Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
> rate 0bit 0pps droprate 0bit backlog 0b 0p requeues 0
> qdisc sfq 120: dev eth0 parent 1:120 limit 128p quantum 1514b flows
> 128/1024 perturb 10sec
> Sent 22796 bytes 185 pkt (dropped 0, overlimits 0 requeues 0)
> rate 0bit 0pps droprate 0bit backlog 0b 0p requeues 0
> qdisc pfifo_fast 0: dev ra0 root bands 3 priomap 1 2 2 2 1 2 0 0 1 1
> 1 1 1 1 1 1
> Sent 19767464443 bytes 29505220 pkt (dropped 0, overlimits 0 requeues 0)
> rate 0bit 0pps droprate 0bit backlog 0b 0p requeues 0
> qdisc pfifo 8009: dev eth0.3900 limit 1p
> Sent 6661359886 bytes 9028157 pkt (dropped 0, overlimits 0 requeues 0)
> rate 154128bit 14pps droprate 0bit backlog 0b 0p requeues 0
> #
>
> eth0 = LAN interface
> ra0 = WLAN interface
> eth0.3900 = Ethernet WAN interface
>
>
>
> Its got an old kernel, with lots of patches to kernel I think, so not
> sure if its easily done to create a new kernel for it either.
>
> # uname -an
> Linux P-2612HNU-F3 2.6.20.19 #44 Wed Mar 30 10:58:43 CST 2011 mips unknown
> # cat /proc/cpuinfo
> system type : AR9
> processor : 0
> cpu model : MIPS 34K V4.12
> BogoMIPS : 261.32
> wait instruction : yes
> microsecond timers : yes
> tlb_entries : 16
> extra interrupt vector : yes
> hardware watchpoint : yes
> ASEs implemented : mips16 dsp mt
> VCED exceptions : not available
> VCEI exceptions : not available
>
> # cat /proc/meminfo
> MemTotal: 124096 kB
> MemFree: 27316 kB
> Buffers: 7512 kB
> Cached: 64772 kB
> SwapCached: 0 kB
> Active: 44640 kB
> Inactive: 37276 kB
> SwapTotal: 0 kB
> SwapFree: 0 kB
> Dirty: 0 kB
> Writeback: 0 kB
> AnonPages: 9640 kB
> Mapped: 27800 kB
> Slab: 9620 kB
> SReclaimable: 1388 kB
> SUnreclaim: 8232 kB
> PageTables: 760 kB
> NFS_Unstable: 0 kB
> Bounce: 0 kB
> CommitLimit: 62048 kB
> Committed_AS: 43156 kB
> VmallocTotal: 1048404 kB
> VmallocUsed: 4164 kB
> VmallocChunk: 1043016 kB
> #
>
>
>
> Pedro
>
> On Wed, May 11, 2011 at 1:22 AM, Dave Taht <dave.taht@gmail.com> wrote:
> > I did a preliminary test of a bismark router with ECN and red enabled
> using
> > a lightly modified
> > version of openwrt's (nbd's) qos-scripts, doing a scp upload from my
> laptop
> > to a server at bufferbloat.net.
> >
> > I artificially throttled the uplink at 1Mbit and downlink to 10Mbit (the
> > measured real values are 24Mbit and 67Mbit respectively) for this test.
> >
> > (there are a few other qdiscs in use, too)
> >
> >
> > I won't claim to have configured red properly but...
> >
> > http://www.teklibre.com/~d/bismarkredc2d_owin.png
> >
> > Look better than usual bufferbloat to y'all?
> >
> > This was over ipv4. Thus far I've been unable to tunnel ipv6 out of
> gatech.
> > There is native
> > IPv6 around here... somewhere....
> >
> > --
> > Dave Täht
> > SKYPE: davetaht
> > US Tel: 1-239-829-5608
> > http://the-edge.blogspot.com
> >
> > _______________________________________________
> > Bloat mailing list
> > Bloat@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/bloat
> >
> >
>
>
>
> --
> Best regards / Mvh
> Jan Pedro Tumusok
>
--
Dave Täht
SKYPE: davetaht
US Tel: 1-239-829-5608
http://the-edge.blogspot.com
[-- Attachment #2: Type: text/html, Size: 6389 bytes --]
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [Bloat] Bismark router + ecn + red tested
2011-05-11 8:49 ` Pedro Tumusok
2011-05-11 13:35 ` Dave Taht
@ 2011-05-11 18:42 ` Jesper Dangaard Brouer
1 sibling, 0 replies; 4+ messages in thread
From: Jesper Dangaard Brouer @ 2011-05-11 18:42 UTC (permalink / raw)
To: Pedro Tumusok; +Cc: bloat
On Wed, 2011-05-11 at 10:49 +0200, Pedro Tumusok wrote:
> I got a Linux based router, ZyXEL P2612 (ADSL2+) and also the P2812
> (VDSL2) both got an ethernet wan port also.
> I can mess around with qdisc on it, since the web ui QoS setup just
> creates qdisc settings to correspond.
Please remember to take the ADSL overhead into account, when configuring
HTB, or else we/you don't have control over the queue, which is the
hole-point when trying to solve bufferbloat by AQM/shaping.
Simply add parameters when creating qdisc:
linklayer adsl
and
overhead XXX
BTW, I don't think its needed, when using VDSL2, as I think its
linklayer is not ATM.
> [...] cut [...]
> Its got an old kernel, with lots of patches to kernel I think, so not
> sure if its easily done to create a new kernel for it either.
>
> # uname -an
> Linux P-2612HNU-F3 2.6.20.19 #44 Wed Mar 30 10:58:43 CST 2011 mips unknown
Damm, linklayer overhead got introduced in 2.6.24.
--
Best regards,
Jesper Dangaard Brouer
ComX Networks A/S
Linux Network Kernel Developer
Cand. Scient Datalog / MSc.CS
Author of http://adsl-optimizer.dk
LinkedIn: http://www.linkedin.com/in/brouer
^ permalink raw reply [flat|nested] 4+ messages in thread
