less that 1gb of transfer wired routers [SOLVED]

Post new topic   This topic is locked: you cannot edit posts or make replies.    DD-WRT Forum Index -> Atheros WiSOC based Hardware
Goto page 1, 2, 3  Next
Author Message
royitoroy
DD-WRT User


Joined: 07 Jun 2007
Posts: 244
Location: La Paz, Bolivia

PostPosted: Fri Sep 03, 2021 23:38    Post subject: less that 1gb of transfer wired routers [SOLVED] Reply with quote
I had never paid attention to the speed of my network because I thought it was very good for what I was doing, but nevertheless I always suffered transferring 10gb or 20gb files from my data server to my computer until I get tired and started investigating And here at the forum helped me to realize that the cable from my server to my router was bad, now they have changed it and transmit to 1gb and now becoming my fastest device, thanks for all who helped me.

But I have noticed that my network is very slow and that my routers connected by cable do not even reach 400mbps of transmission:

Step to show current tests

MAINROUTER DD-WRT v3.0-r47206 std (08/19/21) 192.168.2.1
NETGEAR R9000 1gb port lan x 6 & 10gb port lan x 1
, ,

1ST WIRED AP (from lan to wan port) DD-WRT v3.0-r47225 std (08/23/21) 192.168.2.2
Archer C7v2 1gb port lan x 4
, ,

Code:
root@MAINROUTER_DD-WRT_R9000:~# iperf3 -c 192.168.2.2 -V -P20



_________________

Fiber Modem/Router: ZTE-ZXHN F670L ►►►►►► Internet 1
2G,3G,4G Modem: Amplimax FIT Elsys EPRL18 ►► Internet 2
(failover)

MAINROUTER:
Raspberry PI 4 ►► OpenWrt 19.07.8

Load Balancer, Failover, Vlans, QoS, DHCP.

AP 1:
NETGEAR NIGTHAWK R9000 ►► DD-WRT v3.0-r47510 std

Vanilla, 2.4Ghz, 5Ghz, WPA2, Sharing USB Printer, Wired OpenMediaVault Data Server.
AP 2:
TP-LINK ARCHER AC1750 C7 v2 ►► DD-WRT v3.0-r47225 std
AP 3 & 4
TP-LINK ARCHER AC1750 A7 v5.6 ►► DD-WRT v3.0-r47282 std


Tutorials & WikisDownload FirmwaresBest Wifi SettingsComplete Guide Wifi
Sponsor
kernel-panic69
DD-WRT Guru


Joined: 08 May 2018
Posts: 14126
Location: Texas, USA

PostPosted: Sat Sep 04, 2021 0:42    Post subject: Reply with quote
You will not get any better throughput on your hardware. You can try cubic TCP congestion control across the board and it might improve slightly, but I wouldn't hold my breath.

https://deviwiki.com/wiki/TP-LINK_Archer_C7_v5.x

https://wikidevi.wi-cat.ru/TP-LINK_Archer_A7_v5.x

kernel-panic69 wrote:
Also could be a limitation of the antique hardware or hardware issue in the server, or a fudged config in Debian 10.

https://ark.intel.com/content/www/us/en/ark/products/36461/intel-desktop-board-dp43tf.html

https://ark.intel.com/content/www/us/en/ark/products/33925/intel-pentium-processor-e2200-1m-cache-2-20-ghz-800-mhz-fsb.html

https://en.wikipedia.org/wiki/LGA_775


Wildlion wrote:
I get what people are saying, but I have an older machine with an E6600 Core 2 duo and it can do ~400 Mbps max and it has negotiated a 1 Gbps link through the motherboard, and it is/has run Debian 5,6,7,8,9,10,11,12

_________________
"Life is but a fleeting moment, a vapor that vanishes quickly; All is vanity"
Contribute To DD-WRT
Pogo - A minimal level of ability is expected and needed...
DD-WRT Releases 2023 (PolitePol)
DD-WRT Releases 2023 (RSS Everything)

----------------------
Linux User #377467 counter.li.org / linuxcounter.net
royitoroy
DD-WRT User


Joined: 07 Jun 2007
Posts: 244
Location: La Paz, Bolivia

PostPosted: Sat Sep 04, 2021 2:23    Post subject: Reply with quote
But my dubth with my network is about the transmission speed between the same routers wired to each other, if all my routers have 1gb ports

The tests are between themselves, not with devices connected to them, because I am already clear that whether the Wi-Fi connection depends on various aspects and variables of the client devices, but the test is between the routers themselves.

When it is connection WDS says that the speed drops to half for clients connected to Wi-Fi, that was the 1st reason that i preferred a connection in AP mode so that the loss to the client was not so much, however the wired connection between the R9000 and The C7 like the other AP's does not even reach 300mbs.

Could it be some aspect of the configuration itself? Or something else?

_________________

Fiber Modem/Router: ZTE-ZXHN F670L ►►►►►► Internet 1
2G,3G,4G Modem: Amplimax FIT Elsys EPRL18 ►► Internet 2
(failover)

MAINROUTER:
Raspberry PI 4 ►► OpenWrt 19.07.8

Load Balancer, Failover, Vlans, QoS, DHCP.

AP 1:
NETGEAR NIGTHAWK R9000 ►► DD-WRT v3.0-r47510 std

Vanilla, 2.4Ghz, 5Ghz, WPA2, Sharing USB Printer, Wired OpenMediaVault Data Server.
AP 2:
TP-LINK ARCHER AC1750 C7 v2 ►► DD-WRT v3.0-r47225 std
AP 3 & 4
TP-LINK ARCHER AC1750 A7 v5.6 ►► DD-WRT v3.0-r47282 std


Tutorials & WikisDownload FirmwaresBest Wifi SettingsComplete Guide Wifi
ho1Aetoo
DD-WRT Guru


Joined: 19 Feb 2019
Posts: 2927
Location: Germany

PostPosted: Sat Sep 04, 2021 6:23    Post subject: Reply with quote
if the Archer is configured in "Router Mode", it acts as a swtich, everything that goes over the LAN cables to the clients is only switched.

Regardless of the CPU performance of the router itself, the client should then receive 1Gbit.

R9000 <--- LAN ----> Archer Switch <-- LAN --> Client

if you run Iperf directly on the router CPU then you have to keep in mind that the Archer has only a single core processor with ~700Mhz.

while the R9000 has 4x1,7Ghz

a Gigabit Speed test on my R7800 with 2x 1,7Ghz needs about 60-70% CPU

iperf3 -Rc 192.168.1.1 -P 8 -t 60

CPU: 1.3% usr 26.3% sys 0.0% nic 32.1% idle 0.0% io 0.0% irq 40.1% sirq

so check the CPU load on the routers while the test is running ...

in your picture you can also see 83,2% CPU Load on the remote side at ~245Mbit

royitoroy wrote:



test with less threads - less threads is less CPU load

Edit: besides, this seems to me to be a lot of retransmissions in 10sec (5357)
and as KP already pointed out (I use Cubic myself)

a test between my raspberry pi and my r7800


Last edited by ho1Aetoo on Sat Sep 04, 2021 10:16; edited 3 times in total
Per Yngve Berg
DD-WRT Guru


Joined: 13 Aug 2013
Posts: 6858
Location: Romerike, Norway

PostPosted: Sat Sep 04, 2021 9:21    Post subject: Reply with quote
You have a mis-configuration on the R9000.

The local lan IP address and gateway are not on the same sub-net (192.168.2.x).
royitoroy
DD-WRT User


Joined: 07 Jun 2007
Posts: 244
Location: La Paz, Bolivia

PostPosted: Sat Sep 04, 2021 13:06    Post subject: Reply with quote
Per Yngve Berg wrote:
You have a mis-configuration on the R9000.

The local lan IP address and gateway are not on the same sub-net (192.168.2.x).


are you mean this?



My internet modem ZTE (which I can't configure at all because I don't have the credentials) is 192.168.1.1 and it gives me a private ip 192.168.1.200.

before I had it empty and then put the mainrouter's lan ip 192.168.1.1 but it stop Of having internet in the 2 public VAP that i have (or WAP in AP2 & AP4)

What ip should i put there? Because we fought a lot so that the VAP had internet in the first place

_________________

Fiber Modem/Router: ZTE-ZXHN F670L ►►►►►► Internet 1
2G,3G,4G Modem: Amplimax FIT Elsys EPRL18 ►► Internet 2
(failover)

MAINROUTER:
Raspberry PI 4 ►► OpenWrt 19.07.8

Load Balancer, Failover, Vlans, QoS, DHCP.

AP 1:
NETGEAR NIGTHAWK R9000 ►► DD-WRT v3.0-r47510 std

Vanilla, 2.4Ghz, 5Ghz, WPA2, Sharing USB Printer, Wired OpenMediaVault Data Server.
AP 2:
TP-LINK ARCHER AC1750 C7 v2 ►► DD-WRT v3.0-r47225 std
AP 3 & 4
TP-LINK ARCHER AC1750 A7 v5.6 ►► DD-WRT v3.0-r47282 std


Tutorials & WikisDownload FirmwaresBest Wifi SettingsComplete Guide Wifi
royitoroy
DD-WRT User


Joined: 07 Jun 2007
Posts: 244
Location: La Paz, Bolivia

PostPosted: Sat Sep 04, 2021 13:23    Post subject: Reply with quote
ho1Aetoo wrote:
if the Archer is configured in "Router Mode", it acts as a swtich, everything that goes over the LAN cables to the clients is only switched.

Regardless of the CPU performance of the router itself, the client should then receive 1Gbit.

R9000 <--- LAN ----> Archer Switch <-- LAN --> Client


Exactly that is what should happen

,


Quote:
if you run Iperf directly on the router CPU then you have to keep in mind that the Archer has only a single core processor with ~700Mhz.

while the R9000 has 4x1,7Ghz

a Gigabit Speed test on my R7800 with 2x 1,7Ghz needs about 60-70% CPU

iperf3 -Rc 192.168.1.1 -P 8 -t 60

CPU: 1.3% usr 26.3% sys 0.0% nic 32.1% idle 0.0% io 0.0% irq 40.1% sirq


ok and where could i see that data as well cpu, usr, sys etc because i just have this and dont kno if its so truthful



Quote:
so check the CPU load on the routers while the test is running...

in your picture you can also see 83,2% CPU Load on the remote side at ~245Mbit

royitoroy wrote:



test with less threads - less threads is less CPU load

Edit: besides, this seems to me to be a lot of retransmissions in 10sec (5357)
and as KP already pointed out (I use Cubic myself)


ok i will do yor same ipref3 code just withna vervos added 1st en westwood next with cubic and give notice as well

_________________

Fiber Modem/Router: ZTE-ZXHN F670L ►►►►►► Internet 1
2G,3G,4G Modem: Amplimax FIT Elsys EPRL18 ►► Internet 2
(failover)

MAINROUTER:
Raspberry PI 4 ►► OpenWrt 19.07.8

Load Balancer, Failover, Vlans, QoS, DHCP.

AP 1:
NETGEAR NIGTHAWK R9000 ►► DD-WRT v3.0-r47510 std

Vanilla, 2.4Ghz, 5Ghz, WPA2, Sharing USB Printer, Wired OpenMediaVault Data Server.
AP 2:
TP-LINK ARCHER AC1750 C7 v2 ►► DD-WRT v3.0-r47225 std
AP 3 & 4
TP-LINK ARCHER AC1750 A7 v5.6 ►► DD-WRT v3.0-r47282 std


Tutorials & WikisDownload FirmwaresBest Wifi SettingsComplete Guide Wifi
royitoroy
DD-WRT User


Joined: 07 Jun 2007
Posts: 244
Location: La Paz, Bolivia

PostPosted: Sat Sep 04, 2021 14:04    Post subject: Reply with quote
westwood test 1
no started

during the test
result 1

cubic test 2
no started

during test
result 2

Quote:
Note: during test 1 the cpu graph always remained below 70 (normally it is at 50
56 with peaks of 63
64%) and only in the last 7 seconds at the time of adding the entire result did it go up to 80
90%.

In test 2 almost during the whole test I never go below 80
90% reaching 100% at times


EXTRA test 3 reno (I read that it is the predecessor of westwood and in some tests it showed to be better just to discart)

no started

during test
result 3


whats do you think? the three test ar ligthly the same...

reno seem to be ligthly more efficient in cpu... but in none of three reach at least 500mbps bteetween them

_________________

Fiber Modem/Router: ZTE-ZXHN F670L ►►►►►► Internet 1
2G,3G,4G Modem: Amplimax FIT Elsys EPRL18 ►► Internet 2
(failover)

MAINROUTER:
Raspberry PI 4 ►► OpenWrt 19.07.8

Load Balancer, Failover, Vlans, QoS, DHCP.

AP 1:
NETGEAR NIGTHAWK R9000 ►► DD-WRT v3.0-r47510 std

Vanilla, 2.4Ghz, 5Ghz, WPA2, Sharing USB Printer, Wired OpenMediaVault Data Server.
AP 2:
TP-LINK ARCHER AC1750 C7 v2 ►► DD-WRT v3.0-r47225 std
AP 3 & 4
TP-LINK ARCHER AC1750 A7 v5.6 ►► DD-WRT v3.0-r47282 std


Tutorials & WikisDownload FirmwaresBest Wifi SettingsComplete Guide Wifi
Alozaros
DD-WRT Guru


Joined: 16 Nov 2015
Posts: 6410
Location: UK, London, just across the river..

PostPosted: Sat Sep 04, 2021 14:32    Post subject: Reply with quote
TCP BBR
Bottleneck Bandwidth and Round-trip propagation time (BBR) is a CCA developed at Google in 2016.[25] While most CCAs are loss-based, in that they rely on packet loss to detect congestion and lower rates of transmission, BBR, like TCP Vegas, is model-based. The algorithm uses the maximum bandwidth and round-trip time at which the network delivered the most recent flight of outbound data packets to build a model of the network. Each cumulative or selective acknowledgment of packet delivery produces a rate sample which records the amount of data delivered over the time interval between the transmission of a data packet and the acknowledgment of that packet.[26] As network interface controllers evolve from megabit per second to gigabit per second performance, the latency associated with bufferbloat instead of packet loss becomes a more reliable marker of the maximum throughput, making model-based CCAs which provide higher throughput and lower latency, such as BBR, a more reliable alternative to more popular loss-based algorithms like TCP CUBIC.

so, for best performance choose either BBR or CUBIC !

to be honest dig deeply as R9000 should deliver much more...than 100Mbit... try all the ports, check cables..be sure ISP is providing whatever is expected...

_________________
Atheros
TP-Link WR740Nv1 ---DD-WRT 55179 WAP
TP-Link WR1043NDv2 -DD-WRT 55303 Gateway/DoT,Forced DNS,Ad-Block,Firewall,x4VLAN,VPN
TP-Link WR1043NDv2 -Gargoyle OS 1.15.x AP,DNS,QoS,Quotas
Qualcomm-Atheros
Netgear XR500 --DD-WRT 55460 Gateway/DoH,Forced DNS,AP Isolation,4VLAN,Ad-Block,Firewall,Vanilla
Netgear R7800 --DD-WRT 55460 Gateway/DoT,AD-Block,Forced DNS,AP&Net Isolation,x3VLAN,Firewall,Vanilla
Netgear R9000 --DD-WRT 55363 Gateway/DoT,AD-Block,AP Isolation,Firewall,Forced DNS,x2VLAN,Vanilla
Broadcom
Netgear R7000 --DD-WRT 55460 Gateway/SmartDNS/DoH,AD-Block,Firewall,Forced DNS,x3VLAN,VPN
NOT USING 5Ghz ANYWHERE
------------------------------------------------------
Stubby DNS over TLS I DNSCrypt v2 by mac913
kernel-panic69
DD-WRT Guru


Joined: 08 May 2018
Posts: 14126
Location: Texas, USA

PostPosted: Sat Sep 04, 2021 15:27    Post subject: Reply with quote
I doubt those Archers have BBR since you have to have a certain kernel version to have it, unless it is backported. Which is why I said cubic. Also, we're talking lan-side only throughput, Alozaros. Pay attention.
_________________
"Life is but a fleeting moment, a vapor that vanishes quickly; All is vanity"
Contribute To DD-WRT
Pogo - A minimal level of ability is expected and needed...
DD-WRT Releases 2023 (PolitePol)
DD-WRT Releases 2023 (RSS Everything)

----------------------
Linux User #377467 counter.li.org / linuxcounter.net
ho1Aetoo
DD-WRT Guru


Joined: 19 Feb 2019
Posts: 2927
Location: Germany

PostPosted: Sat Sep 04, 2021 15:35    Post subject: Reply with quote
@royitoroy


i did not claim that another TCP congestion control would magically pull 250Mbit more out of the hat

that was only related to the retransmissions


based on your tests:

westwood:
15823 retransmissions in 60sec
253Mbit
86,3% load

cubic:
6673 retransmissions in 60sec
269Mbit
86,6% load

reno:
9541 retransmissions in 60sec
260Mbit
79,1% load


so i would say that cubic has already performed the best


but would still be too many retransmissions for me

Code:
[ 17]   0.00-60.00  sec   835 MBytes   117 Mbits/sec                  receiver
[ 19]   0.00-60.00  sec   836 MBytes   117 Mbits/sec    0             sender
[ 19]   0.00-60.00  sec   835 MBytes   117 Mbits/sec                  receiver
[SUM]   0.00-60.00  sec  6.53 GBytes   935 Mbits/sec    0             sender
[SUM]   0.00-60.00  sec  6.52 GBytes   934 Mbits/sec                  receiver


Another bad cable ? Smile

(or possibly too much CPU load)

Edit: and the CPU graph in WebIF is inaccurate, it shows the load average but not the actual CPU load

run the test for a longer time e.g. -t 600 then you will see how the CPU load goes up in WebIF


Last edited by ho1Aetoo on Sat Sep 04, 2021 16:01; edited 2 times in total
kernel-panic69
DD-WRT Guru


Joined: 08 May 2018
Posts: 14126
Location: Texas, USA

PostPosted: Sat Sep 04, 2021 15:59    Post subject: Reply with quote
~250Mbit/s is probably the max that will be seen here. IDGAF what you do. Perhaps a test between R9000 and "server" would see ~1Gbit/s data transfer, but not from the Archers to any other point in the network. I'm out of this circular discussion.
_________________
"Life is but a fleeting moment, a vapor that vanishes quickly; All is vanity"
Contribute To DD-WRT
Pogo - A minimal level of ability is expected and needed...
DD-WRT Releases 2023 (PolitePol)
DD-WRT Releases 2023 (RSS Everything)

----------------------
Linux User #377467 counter.li.org / linuxcounter.net


Last edited by kernel-panic69 on Sat Sep 04, 2021 18:15; edited 1 time in total
ho1Aetoo
DD-WRT Guru


Joined: 19 Feb 2019
Posts: 2927
Location: Germany

PostPosted: Sat Sep 04, 2021 16:24    Post subject: Reply with quote
royitoroy wrote:

ok and where could i see that data as well cpu, usr, sys etc because i just have this and dont kno if its so truthful


Code:
top


or

Code:
htop
ho1Aetoo
DD-WRT Guru


Joined: 19 Feb 2019
Posts: 2927
Location: Germany

PostPosted: Sat Sep 04, 2021 16:41    Post subject: Reply with quote
kernel-panic69 wrote:
~250Mbit/s is probably the max that will be seen here. IDGAF what you do. I don't think even unmanaged switches in place of the A7s and C7 and a test between R9000 and "server" would see a full 1Gbit/s data transfer. The link speed is irrelevant; I think @royitoroy seems to think that gigabit ethernet means you get 1Gbit/s data transfer, no matter what. I'm out of this circular discussion.


he tests between the R9000 and the Archer

that is so slow because of the slow CPU of the Archer

he should test between the R9000 and the NAS - that should be much faster, depending on the switch performance to 900Mbit

on NAS
Code:
iperf3 -c 192.168.2.1 -P8 -t60


R9000 <-- LAN --> Archer Switch <-- LAN --> NAS
Per Yngve Berg
DD-WRT Guru


Joined: 13 Aug 2013
Posts: 6858
Location: Romerike, Norway

PostPosted: Sat Sep 04, 2021 16:50    Post subject: Reply with quote
royitoroy wrote:
My internet modem ZTE (which I can't configure at all because I don't have the credentials) is 192.168.1.1 and it gives me a private ip 192.168.1.200.

before I had it empty and then put the mainrouter's lan ip 192.168.1.1 but it stop Of having internet in the 2 public VAP that i have (or WAP in AP2 & AP4)

What ip should i put there? Because we fought a lot so that the VAP had internet in the first place


It's the gateway for the lan interface. Your modem is not connected to a lan port. Leave it blank.

If you have another router, put it's IP at 192.168.2.x here.
Goto page 1, 2, 3  Next Display posts from previous:    Page 1 of 3
Post new topic   This topic is locked: you cannot edit posts or make replies.    DD-WRT Forum Index -> Atheros WiSOC based Hardware All times are GMT

Navigation

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You cannot download files in this forum