hello,
i run this test which indicate that my network isn't ideal for online gaming and video conferencing
the resault is here:
i gets ping around 60ms to 110ms, any help would apprecicated thanks
Buy more speed!
You only have up to 2Mbps.
I recommend at least 6Mbps.
1 Like
i have 30mbps but i limited the speed because i used 75% of my mounthly quota
the result after i limited speed to up 2800kbit and down 18000kbit
People would probably find it easier to help if you actually told us these basic facts to your question to begin with?
But we still need to know how you limited the speed?
And what actual ISP connection do you have (fiber/dsl…)?
But doest it matter? You want to play games and have video calls but at the same time has so low surf pot that you need to limit the speed to survive the speed test?
I don’t care about your manually limited speed because that means you have a overhead processing of data that scrambles every measurement you do.
What is the speed at fully released trottles?
2 Likes
i have vdsl connection that bridged from modem to openwrt
and my wan interface is pppoe-wan, i limited speed by this script https://raw.githubusercontent.com/Last-times/CAKE-QoS-Script-OpenWrt/main/cake.sh
and fully internet speed are as follow without any traffic shaper:
DOWNLOAD Mbps
22.26
UPLOAD Mbps
2.88
and modem is zte h168n it support tc qdisc and linux commands and openwrt device is x86_64 with one ethernet i setup vlan eth0.1 for lan port and eth0.4094 for wan port
This test is based on both the absolute/unloaded latency and the increases caused by up- or download saturation. Both of your tests indicate that the unloaded latency is your main issue, you are simply 80-90ms away from the next data center housing the servers the test used. 80ms latency/RTT translates roughly to a distance of 8000km (there and back signal propagation delay in optic fiber or copper wires).
So just based on those tests there is nothing you need or even could do, besides moving closer to the test servers (which would improve your test scores, but move you further away from your home 
).
Just out of curiosity, what is the output of tc -s qdisc?
2 Likes
the output of tc -s qdisc without script:
qdisc noqueue 0: dev lo root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc fq_codel 0: dev eth0 root refcnt 2 limit 10240p flows 1024 quantum 1514 target 5ms interval 100ms memory_limit 32Mb ecn drop_batch 64
Sent 69249162 bytes 88361 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
maxpacket 1446 drop_overlimit 0 new_flow_count 3 ecn_mark 0
new_flows_len 0 old_flows_len 0
qdisc noqueue 0: dev br-lan root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc noqueue 0: dev eth0.1 root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc noqueue 0: dev eth0.4094 root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc fq_codel 0: dev pppoe-wan root refcnt 2 limit 10240p flows 1024 quantum 1514 target 5ms interval 100ms memory_limit 32Mb ecn drop_batch 64
Sent 10074568 bytes 36758 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
maxpacket 136 drop_overlimit 0 new_flow_count 5 ecn_mark 0
new_flows_len 0 old_flows_len 0
with script:
qdisc noqueue 0: dev lo root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc fq_codel 0: dev eth0 root refcnt 2 limit 10240p flows 1024 quantum 1514 target 5ms interval 100ms memory_limit 32Mb ecn drop_batch 64
Sent 168007567 bytes 236979 pkt (dropped 0, overlimits 0 requeues 9)
backlog 0b 0p requeues 9
maxpacket 1514 drop_overlimit 0 new_flow_count 101 ecn_mark 0
new_flows_len 0 old_flows_len 0
qdisc noqueue 0: dev lanbrport root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc cake 800f: dev lanveth root refcnt 2 bandwidth 16Mbit diffserv8 dual-dsthost nonat nowash ingress no-ack-filter split-gso rtt 98ms ptm overhead 30
Sent 46771920 bytes 49572 pkt (dropped 16, overlimits 47647 requeues 0)
backlog 0b 0p requeues 0
memory used: 209088b of 4Mb
capacity estimate: 16Mbit
min/max network layer size: 28 / 1492
min/max overhead-adjusted size: 59 / 1546
average network hdr offset: 14
Tin 0 Tin 1 Tin 2 Tin 3 Tin 4 Tin 5 Tin 6 Tin 7
thresh 16Mbit 14Mbit 12250Kbit 10718Kbit 9378Kbit 8206Kbit 7180Kbit 6283Kbit
target 4.9ms 4.9ms 4.9ms 4.9ms 4.9ms 4.9ms 4.9ms 4.9ms
interval 98ms 98ms 98ms 98ms 98ms 98ms 98ms 98ms
pk_delay 0us 559us 3us 2.92ms 0us 0us 514us 155us
av_delay 0us 151us 0us 570us 0us 0us 234us 2us
sp_delay 0us 6us 0us 5us 0us 0us 7us 2us
backlog 0b 0b 0b 0b 0b 0b 0b 0b
pkts 0 8394 5 28117 0 0 13030 42
bytes 0 7809079 1976 31976955 0 0 7002432 1764
way_inds 0 0 0 11 0 0 0 0
way_miss 0 9 2 298 0 0 1 1
way_cols 0 0 0 0 0 0 0 0
drops 0 0 0 16 0 0 0 0
marks 0 0 0 0 0 0 0 0
ack_drop 0 0 0 0 0 0 0 0
sp_flows 0 1 2 0 0 0 1 0
bk_flows 0 0 0 1 0 0 0 0
un_flows 0 0 0 0 0 0 0 0
max_len 0 1506 1007 1506 0 0 1284 42
quantum 488 427 373 327 300 300 300 300
qdisc noqueue 0: dev br-lan root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc noqueue 0: dev eth0.1 root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc noqueue 0: dev eth0.4094 root refcnt 2
Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
qdisc cake 8010: dev pppoe-wan root refcnt 2 bandwidth 2Mbit diffserv8 dual-srchost nat wash no-ack-filter split-gso rtt 98ms ptm overhead 30
Sent 13282374 bytes 43841 pkt (dropped 149, overlimits 43493 requeues 0)
backlog 0b 0p requeues 0
memory used: 153792b of 4Mb
capacity estimate: 2Mbit
min/max network layer size: 40 / 1492
min/max overhead-adjusted size: 72 / 1546
average network hdr offset: 0
Tin 0 Tin 1 Tin 2 Tin 3 Tin 4 Tin 5 Tin 6 Tin 7
thresh 2Mbit 1750Kbit 1531Kbit 1339Kbit 1172Kbit 1025Kbit 897576bit 785376bit
target 9.08ms 10.4ms 11.9ms 13.6ms 15.5ms 17.7ms 20.2ms 23.1ms
interval 102ms 103ms 105ms 107ms 109ms 111ms 113ms 116ms
pk_delay 0us 10.4ms 1.17ms 14.3ms 90us 0us 1.09ms 3.29ms
av_delay 0us 1.62ms 341us 2.21ms 1us 0us 36us 176us
sp_delay 0us 14us 7us 21us 1us 0us 9us 4us
backlog 0b 0b 0b 0b 0b 0b 0b 0b
pkts 0 9574 1648 20230 8 0 12459 71
bytes 0 7648988 120717 4298615 608 0 1405622 16076
way_inds 0 0 16 0 0 0 0 0
way_miss 0 9 1375 303 8 0 2 3
way_cols 0 0 0 0 0 0 0 0
drops 0 103 0 46 0 0 0 0
marks 0 0 0 2 0 0 0 0
ack_drop 0 0 0 0 0 0 0 0
sp_flows 0 1 32 0 1 0 0 0
bk_flows 0 0 0 1 0 0 0 0
un_flows 0 0 0 0 0 0 0 0
max_len 0 2173 301 2473 76 0 229 445
quantum 300 300 300 300 300 300 300 300
Thanks! At low rates (like you upload's 2Mbps) it might not make too much sense to use difserv8 as some of the tins only get relatively little guaranteed traffic rates (e.g. your upload on pppoe-wan, tin 7 only gets 785Kbps)... yes if traffic exceeds its tin it will not be dropped put into one of the lower tins (not sure which, or if whole flows are demoted) but this counteracts the whole purpose of putting flows into higher priority tins in the first place.
1 Like
This is based on @elan script, However i seen also he mentioned that diffserv8 is broken however someone is using diffserv8 and seems fine. Qosify: new package for DSCP marking + cake - #362 by EnfermeraSexy
You can’t even run Netflix on this. Video conference with outgoing video stream will probably lag a lot for all the others watching you or be auto set for low resolution under up/download speed of 6Mbps.
In upload direction? Did I miss that he wanted to serve Netflix from his home?
Really? Zoom claims:
Group video call in 480p SD: 800 Kbps (0.8 Mbps)/1.0 Mbps
Group call/gallery view in 720p HD: 1.5 Mbps/1.5 Mbps
Sending and receiving group call video in 1080p HD: 2.5 Mbps/3.0 Mbps
I agree that for 1080p this will not work, but 480 sure should and 720 might.
1 Like
That is exactly what I said. The video quality must go down a lot or only sound to work with that upload speed.
Sorry, I misunderstood you then. IMHO 480p video is plenty for most work VCs, that is the content does not require/merit higher resolutions ;)... but it is certainly helpful to know the limits of what is possible...
I can confirm ~1.45 Mbps upload handles my work VCs and VOIP audio calls just fine.
I have two ISP options: 230/12 Mbps cable and approximately 26/1.6 Mbps bonded VDSL. When the low cost introductory period (usually ~12 months) ends with one ISP, I switch to the other. I'm in a slow VDSL period presently 
Unloaded latency on my slow VDSL option is ~36 ms. Without SQM, loading adds ~125 ms to download and ~1100 ms to upload, so basically unusable for most everything and a well deserved "F" rating. Layer CAKE SQM reduces my 1.6 Mbps upload to ~1.45 Mbps, but with SQM loading adds only ~4-8 ms down and up to the unloaded ~36 ms, delivering a consistent "A" buffer bloat rating. CAKE is great!
Setting download speed to 23000 and upload to 1550 seems to give me the best consistent low latency. I experiment raising each until I notice some degradation starting to set in. I turn ECN on for both ingress and egress. I also use Ethernet with overhead link layer adaption with per packet overhead set to 34 bytes.
1 Like