[Cake] New to cake. Some questions

Sebastian Moeller moeller0 at gmx.de
Fri Jun 10 10:26:01 EDT 2016 

Hi Dennis,?

On June 10, 2016 4:05:11 PM GMT+02:00, Dennis Fedtke <dennisfedtke at gmail.com> wrote:
>Hi Sebastian,
>
>yes this is wired connection. As i stated my ping times always vary 
>independently of target.
>My ISP is overloaded in certain regions. So i assume they do some 
>shaping/limiting on certain protocols (icmp for example)
>Connection speed is 200/20 Mbit.

Okay that is a Docsis cable Link, so no atm encapsulation at all. There a several lines of reasoning to one to this conclusion, but mainly ATM links top out at 22Mbps, and in Germany typically at 16-17Mbps, and your ISP is a pure cable. Company. So you can stop running the overhead detector as that only works on ATM links, sorry. I have not yet found a way to measure the overhead without ATM cells.

>ISP is unitymedia which doesn't allow you to use your own hardware.

The law changed an soon, August I believe they will have to give you the access information, but you will still need a cable modem or Docsis router...


>So actually i have to run my router behind theirs with exposed host 
>enabled :<
>
>Ping response:
>
>ping -s 1400 -c 1 109.90.x.x
>PING 109.90.28.1 (109.90.28.1) 1400(1428) bytes of data.
>1408 bytes from 109.90.28.1: icmp_seq=1 ttl=253 time=11.6 ms
>
>--- 109.90.28.1 ping statistics ---
>1 packets transmitted, 1 received, 0% packet loss, time 0ms
>rtt min/avg/max/mdev = 11.677/11.677/11.677/0.000 ms
>
>This looks good or?

      Yes the ping is fine, I assume that your node is quite overbooked and you the 4ms variance from the Docsis grant request system or so. But I am no Docsis expert, so that could be wrong...


>
>Yes i am from germany. So you are from germany too?

       Yes.

>
>Thanks for your time and help :)

        Happy to be able to help....
>
>
>Best regards
>Dennis

Freundlichem Gruessen
        Sebastian
>
>Am 10.06.2016 um 15:02 schrieb moeller0:
>> Hi Dennis,
>>
>>> On Jun 10, 2016, at 14:43 , Dennis Fedtke <dennisfedtke at gmail.com>
>wrote:
>>>
>>> Hi Sebastian,
>>>
>>> i used the default setting of 1000.
>> 	Okay, that should work i assume unless you have a very fast link…
>What link at what ISP do you actually have?
>>
>>> But it seems that my isp is dropping icmp packets if there are
>exceeding some sending threshold.
>> 	I would be amazed if they did, a sympotom of that would be rsate
>reduction to all ICMP probe flows independent of target host. If
>however you only see this with specific hosts it is very likely that
>that host rate limits its ICMP responses. In either case try another
>host further upstream. II think I has reasonable decent results with
>targeting 8.8.8.8, googles dns servers.
>>
>>> So there is a lot of none usable ping data.
>> 	Again, try another host…
>>
>>> I increased the send delay to 50ms. 25 ms already shows dropped
>requests.
>> 	That might also help, as long as you stay below their throttling
>rate the chosen host might still work okay.
>>
>>> This is the third run now. Waiting for completion.
>> 	Well, sorry that the method is not as slick and streamlined, but
>there are no guarded good ICMP reflectors available on the net.
>>
>>> The ping target is my first hop.
>> 	Try the next hop then ;)
>>
>>> Actually my ping always varies around +-5ms even at idle and
>independently of ping target.
>> 	This is via wifi/wlan? If so try from a wired connection instead.
>>
>>> When i look through the ping file the increase in ping times are
>actually appear to be random to me.
>> 	Well, we expect variability of the individual “trials” to exist,
>that is why we collect so many and try to select the best measure in
>the matlab code to remove the unwanted variance. Could you post a link
>to both of the generated plots please, the first one showing te
>different aggregation measures might be helpful in diagnosing the
>issues deeper.
>>
>>> So how to test if my isp responses with fixed icmp packet size?
>> 	You could try manually. In the folloewing example I pinged
>gstatic.com (which belongs to googles CDN as far as I know):
>>
>> bash-3.2$ ping -s 1 -c 1 gstatic.com
>> PING gstatic.com (216.58.213.195): 1 data bytes
>> 9 bytes from 216.58.213.195: icmp_seq=0 ttl=55
>>
>> --- gstatic.com ping statistics ---
>> 1 packets transmitted, 1 packets received, 0.0% packet loss
>>
>>
>> bash-3.2$ ping -s 64 -c 1 gstatic.com
>> PING gstatic.com (216.58.213.195): 64 data bytes
>> 72 bytes from 216.58.213.195: icmp_seq=0 ttl=55 time=19.446 ms
>>
>> --- gstatic.com ping statistics ---
>> 1 packets transmitted, 1 packets received, 0.0% packet loss
>> round-trip min/avg/max/stddev = 19.446/19.446/19.446/0.000 ms
>>
>>
>> bash-3.2$ ping -s 65 -c 1 gstatic.com
>> PING gstatic.com (216.58.213.195): 65 data bytes
>> 72 bytes from 216.58.213.195: icmp_seq=0 ttl=55 time=21.138 ms
>> wrong total length 92 instead of 93
>>
>> --- gstatic.com ping statistics ---
>> 1 packets transmitted, 1 packets received, 0.0% packet loss
>> round-trip min/avg/max/stddev = 21.138/21.138/21.138/0.000 ms
>> bash-3.2$
>>
>>
>> bash-3.2$ ping -s 1400 -c 1 gstatic.com
>> PING gstatic.com (216.58.213.195): 1400 data bytes
>> 72 bytes from 216.58.213.195: icmp_seq=0 ttl=55 time=6.878 ms
>> wrong total length 92 instead of 1428
>>
>> --- gstatic.com ping statistics ---
>> 1 packets transmitted, 1 packets received, 0.0% packet loss
>> round-trip min/avg/max/stddev = 6.878/6.878/6.878/0.000 ms
>>
>> Once I try to send 65 Bytes of ICMP payload the response is cut short
>to 92 bytes, the same might happen with your isp. But also if all your
>ISP does is rate limiting the ICMP packests that still can lead to to
>much variance in the RTTs…
>>
>>
>>> Im in central europe too :D
>> 	Ah, then you just have a different work/sleep cycle than I do ;).
>Where in central Europe, if I might as Ii am, as you might have guessed
>based in Germany…
>>
>> Best Regards
>> 	Sebastian
>>
>>> Thanks :)
>>>
>>>
>>> Am 10.06.2016 um 07:20 schrieb moeller0:
>>>> Hi Dennis,
>>>>
>>>>
>>>>> On Jun 10, 2016, at 02:49 , Dennis Fedtke <dennisfedtke at gmail.com>
>wrote:
>>>>>
>>>>> Hi Sebastian,
>>>>>
>>>>> Sorry this is positive or?
>>>> 	I would say that is unclear…
>>>>
>>>>> But i need more samples ?
>>>> 	I would try with more samples, after checking that the ping times
>in the recorded data file actually are larger for larger probes than
>for smaller, some hosts will reply with a fixed maximum ICMP packet
>instead of returning the received packet, thereby reducing the signal
>range (as only the upload leg of the link is meaning fully contributing
>useful differential signal.
>>>> 	BTW I am in central europe so at times of the day my responses can
>be very sporadic, as I either am at work or sleeping ;)
>>>>
>>>> Best Regards
>>>> 	Sebastian
>>>>
>>>>> Thanks :)
>>>>>
>>>>>
>>>>> Am 10.06.2016 um 01:11 schrieb moeller0:
>>>>>> Hi Dennis,
>>>>>>
>>>>>>> On Jun 10, 2016, at 00:45 , Dennis Fedtke
><dennisfedtke at gmail.com> wrote:
>>>>>>>
>>>>>>> Hi Sebastian,
>>>>>>>
>>>>>>> thank you for your answers :)
>>>>>>>
>>>>>>> The ATM overhead detector script is currently running.
>>>>>>> I read the wiki about it but im not quite sure how to interpret
>the plot.
>>>>>>> I mean what info should i read from it? maximum packet size?
>>>>>> 	The relevant number is reported as “Estimated overhead preceding
>the IP header” in the top part of the second figure created by the
>script. But that is only relevant.useful if you see a nice step like
>plot in figure 2 as well ( the second figure in
>https://github.com/moeller0/ATM_overhead_detector/wiki as positive and
>the fourth figure as negative example.
>>>>>>
>>>>>>> If yes do i set the overhead in cake? Or do i set iptables to
>clamp to new mtu/mss?
>>>>>> 	If you use plain cake and you know the numerical overhead (NN)
>the easiest is to add the following to your cake invocation: “atm
>overhead NN”
>>>>>>
>>>>>> Please note that if you use cake on an ethernet interface the
>kernel will already account for 14 byte of ethernet overhead, so if the
>script told you 44 as actual overhead, you use ”overhead 30” to address
>that. If you use a pppoe interface the kernel will most likely not add
>the 14 bytes for you, so then you would use “overhead 44” (I excluded
>the atm option in the last examples for clarity…)
>>>>>>
>>>>>>> Regarding UDP paket dropping problem:
>>>>>>> I just read some forums and users stated that under heavy load
>cake starts to drop udp packets which causes lag ingame.
>>>>>>> My idea was to set ingress/egress to diffserv4 and apply the EF
>dscp mark on those packets.
>>>>>> 	Ell, not a bad idea, but often the problem are in the incoming
>traffic, and unfortunately with the ifb we use we can not use iptables,
>but only tc, and remarking with tc is unpleasant.
>>>>>>
>>>>>>> Will this even work? if yes how to do this? iptables?
>>>>>> 	No, you wuld need tp use tc.
>>>>>>
>>>>>>> ipt -t mangle -A PREROUTING -p udp -m multiport --ports
>5000:5500 -j DSCP --set-dscp-class EF
>>>>>>>
>>>>>>> Like thia? Is prerouting correct here? (Taken from layer cake
>script)
>>>>>> 	This will affect outgoing packets and might be a good idea in
>your specific case.
>>>>>>
>>>>>> BUT why don’t you try the default behaviour with specific rules
>and tricks and report success or failure back to us, after all the
>fastest/easiest classification is one one does not need to perform at
>all.
>>>>>>
>>>>>>> For the squash and wash feature.
>>>>>>> Im asking because if i choose to squash in the advanced options
>of sqm scripts.
>>>>>>> The dscp fields/marks will be overwritten by iptables to 0
>(besteffort). (layer cake script)
>>>>>>> So then it makes no sense to manually set dscp fields/marks or?
>(Or even setting diffserv)
>>>>>> 	No unfortunately on ingress cake sees the packets before
>iptables, so the effective behavioral emulation of wash/squash by cake
>is to set ingress cake to besteffort (basically cake ignores the dscp
>field which functionally is identical to all packets having the same
>value). The squashing by iptables just clears the dscp marls so that
>internal networking elements like potentially wifi liknks are not
>confuzed by the dscp information.
>>>>>>
>>>>>>> Did i understand this correctly. Per rfc isps should not provide
>dscp fields/marks?
>>>>>> 	Not exactly, per RFC DSCPs are only ever valid/defined inside a
>DSCP domain and your ISPs domain ends before it reaches your CPE. Since
>you have no control over your ISPs markings, they can be very much not
>like you want them to be (Dave That reported that his ISP re-mapped
>almost 1/3 or so of packets into the CS1 background class). So it is
>recomended that each DSCP domain re-mapps the code points at its entry
>point, which in your case is your router…
>>>>>>
>>>>>> Best Regards
>>>>>> 	Sebastian
>>>>>>
>>>>>>> Thank you.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Am 09.06.2016 um 23:30 schrieb moeller0:
>>>>>>>> Hi Dennis,
>>>>>>>>
>>>>>>>> let me start with a disclaimer, I am not the best information
>source for cake on this mailing list, but I assume the others will
>chime in if I say something questionable…
>>>>>>>>
>>>>>>>>> On Jun 9, 2016, at 22:58 , Dennis Fedtke
><dennisfedtke at gmail.com> wrote:
>>>>>>>>>
>>>>>>>>> Hi
>>>>>>>>>
>>>>>>>>> Currently im running lede + cake + sqm_scripts and i have some
>questions:
>>>>>>>>>
>>>>>>>>> 1. What is considered the “optimal" setup atm for cake?
>>>>>>>> 	The same as without cake; really, proper per-packet-overhead
>accounting is important for bandwidth shaping, especially for ATM
>-based links. I would recommend to follow the method on
>https://github.com/moeller0/ATM_overhead_detector to m\empirically
>measure whether your link uses ATM encapsulation and what exact
>overhead is in use.
>>>>>>>>
>>>>>>>>> e.g. which cake script should i use piece or layer cake?
>>>>>>>> 	piece_of_cake has only one tier of priority, while layer_cake
>currently offers 4. Packets are put into the different priority bands
>based on the content of their TOS/DSCP filed in the IP header; if this
>is greek to you, I guess piece_of_cake most likely is what you are
>looking for..
>>>>>>>>
>>>>>>>>> 2. Recently squash and wash was removed.
>>>>>>>>> But the sqm scripts were not updated. In the advanced options
>should i set that the dcsp marks are kept?
>>>>>>>> 	This really is an implementation detail that has no immediate
>effect if you choose piece_of_cake as typically only the bottleneck is
>sensitive to DSCP based priority banding. (Typically in that if you are
>unlucky your WLAN will use the DSCP marks to move packets into 4
>different priority classes, which is fine if you want that, but bad for
>not sanity checked packets coming in from the wider internet (one is
>not supposed to assume incoming packets have sensible dscp markings as
>per RFC) that is why the wash/squash option is missed by some of us,
>independent of the fact that it was a layering violation).
>>>>>>>>
>>>>>>>>> 3. Should i use advanced options in sqm scripts and set
>triple-isolate + diffserv8 ?
>>>>>>>> 	If you understand what these options do and believe that this
>is the best for your network go ahead, otherwise… The triple-isolate
>option will try to be fair to host_IP addresses first and then for each
>hostIP fair to each flow, but for that to do something you will most
>likely want this requires that cake sees internal IP addresses of your
>end-hosts. In the typical configuration with SQM on the WAN interface
>of a NAT router all internal addresses are replaced with the external
>IP address of the router it self and triple-isolates per host fairness
>will pretty much be equal to per flow fairness (not exactly, but in
>essence). So if you want to try tiple-isolate or its better defined
>brothers dual-srchost and dual-dsthost you would need to instantiate
>SQM on an internal interface like LAN. But then the direction of
>ingress and egress from the routers perspective changes with regards to
>the internet download and upload direction and you will need to put the
>internet upload bandwidth into the download field of the sqm GUI and
>vice versa. Also SQM on an internal interface will also shape internal
>traffic over the same interface, and that often affects traffic to and
>from the wifi/wlan radios to the lan switch… (I guess you would have
>preferred a shorter less vague response, but such are the constraints…)
>>>>>>>>
>>>>>>>>> 4. Is it recommend to enable diffserv on ingress?
>>>>>>>> 	If you trust/konw/have confirmed that your upstream (ISP?)
>sends you sensible and reasonable DSCP markings by all means enable
>diffserv on ingress. But the default assumption should be that your
>upstream used a dscp mapping that only makes sense for them and not for
>you.
>>>>>>>>
>>>>>>>>> 5. Is there still the udp packet dropping problem? e.g. games
>that are using udp.
>>>>>>>>> If yes does it make sense to apply diffserv classes manually?
>How to do this?
>>>>>>>> 	I am not sure what you mean, but if you test this and have
>some findings please report here…
>>>>>>>>
>>>>>>>>> 6. is the autorate_ingress still under development?
>>>>>>>>> This very interesting feature. especially for docsis networks.
>Will it be possible to set target ping time?
>>>>>>>> 	The last tests did indicate that this feature is not ready for
>primetime at least not on typically fixed bandwidth links and I assume
>docsis links are fixed enough.
>>>>>>>>
>>>>>>>>> 6. What difference does it make to set a different rtt?
>>>>>>>>> Setting lower rtt will reduce download speed i guess but will
>it allow better ping times (because of lower downloadrate uh)?
>>>>>>>>> What happens if rtt is set way higher?
>>>>>>>> 	With the RTT parameter you in essence specify how much time
>you give the endpoints of a flow to respond to a congestion signal (ECN
>marking or packet drop) if you select this way to small you will
>sacrifice bandwidth, if you set this too high you will accumulate more
>latency under load. The good thing seems to be that this does not need
>to be terribly precise, order of magnitude correctness seems to be
>sufficient (at least in base2)
>>>>>>>>
>>>>>>>>
>>>>>>>>> Thank you!
>>>>>>>> 	I am sure the real experts will also chime in…
>>>>>>>>
>>>>>>>> Best Regards
>>>>>>>> 	Sebastian
>>>>>>>>
>>>>>>>>> _______________________________________________
>>>>>>>>> Cake mailing list
>>>>>>>>> Cake at lists.bufferbloat.net
>>>>>>>>> https://lists.bufferbloat.net/listinfo/cake
>>>>>>> _______________________________________________
>>>>>>> Cake mailing list
>>>>>>> Cake at lists.bufferbloat.net
>>>>>>> https://lists.bufferbloat.net/listinfo/cake
>>>

-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

More information about the Cake mailing list