Router can't handle Fiber connection

[Toonces]

I'm running a Linksys WRT54GL with v23 of DD-WRT.
My internet connection is a 100Mbps fiber line with NTT's FLETS here in Japan.
Computer is an ASUS M50VM-A1 with 802.11b/g/n and Gigabit ethernet.

I noticed my speedtest results went down significantly after connecting through the router, both wireless and wired (they maxed at about the same speed). The straight PPPoE from my laptop to the NTT modem was tested against wired through the router using speedtest.net's closest server and finishing the same torrent file through uTorrent; monitoring total download speed and noting the same top peer download speed for both times.

w/ router (wired):
speedtest.net = 9000/8000
torrent = 600kbp/s
top peer = 400kbp/s

PPPoE to modem:
speedtest.net = 32000/14000
torrent = 3.9Mbp/s
top peer = 2.9Mbp/s

*The top peer was another fiber connection here in Japan according to the IP.


My question is:
it looks like WRT54GL can't handle the required throughput of my connection, as when it's part of the network my speeds are drastically reduced.

Anyone have advice as to how this could be resolved? I've heard Tomato firmware increases speeds but enough to overcome this drastic difference?
Otherwise I'm assuming that either replacing the router or buying a gigabit switch would be my two options at this point?

 

[VirtualLarry]

link

Choose a router higher on the list, like a DIR-655.

 

[Engineer]

From here....

OK, I did a methodical but fast trip through more third party firmware. Same as before except I used a better server for iperf (which turned out to not be so much better). I found a strange performance "hole" around 50mbps so redid them all to my main server as before and edited this table. Also, I'm simply reporting the average of the best 5-of-10 measurements.

LAN-to-LAN 85.20
Tomato 1.23 58.42
Linksys 4.71.4.001 50.12
Tomato 1.22 46.84
Tarifa 034RC1 46.40
OpenWrt Kamikaze 7.09 35.46
OpenWrt White Russian 0.9 35.30
DD-WRT 2.4 STD 31.78
DD-WRT 2.4-SP1 STD 31.32
DD-WRT EKO v24-11218_NEWD_std 30.72
DD-WRT EKO v24-11100_NEWD_std 30.72
OpenWrt Kamikaze 8.09RC1 (2.6 kernel) 24.24


The biggest surprise is Tomato 1.23 vs. 1.22. Perhaps someone knows more about this?

Again, keep in mind this is at each firmware's DEFAULT settings. If more features are turned on, performance may suffer.

EDITED with updated results as mentioned above.

 

[Toonces]

Originally posted by: Engineer
From here....

OK, I did a methodical but fast trip through more third party firmware. Same as before except I used a better server for iperf (which turned out to not be so much better). I found a strange performance "hole" around 50mbps so redid them all to my main server as before and edited this table. Also, I'm simply reporting the average of the best 5-of-10 measurements.

LAN-to-LAN 85.20
Tomato 1.23 58.42
Linksys 4.71.4.001 50.12
Tomato 1.22 :beer: 46.84
Tarifa 034RC1 46.40
OpenWrt Kamikaze 7.09 35.46
OpenWrt White Russian 0.9 35.30
DD-WRT 2.4 STD 31.78
DD-WRT 2.4-SP1 STD 31.32
DD-WRT EKO v24-11218_NEWD_std 30.72
DD-WRT EKO v24-11100_NEWD_std 30.72
OpenWrt Kamikaze 8.09RC1 (2.6 kernel) 24.24


The biggest surprise is Tomato 1.23 vs. 1.22. Perhaps someone knows more about this?

Again, keep in mind this is at each firmware's DEFAULT settings. If more features are turned on, performance may suffer.

EDITED with updated results as mentioned above.

:thumbsup::beer:

Thanks Engineer, Tomato looks to have brought my speeds pretty close to what I was getting through direct PPPoE to the modem.

 

[Engineer]

Originally posted by: toonces
:thumbsup::beer:

Thanks Engineer, Tomato looks to have brought my speeds pretty close to what I was getting through direct PPPoE to the modem.

😀

 

[Toonces]

updated scores:

WRT54GL (w/ 1.23 Tomato)
speedtest.net = 26000/12000
torrent = 3.0Mbp/s
top peer = 1.4Mbp/s (not the same IP, so can't make a quantitative comparison)

 

[Engineer]

Originally posted by: toonces
updated scores:

WRT54GL (w/ 1.23 Tomato)
speedtest.net = 26000/12000
torrent = 3.0Mbp/s
top peer = 1.4Mbp/s (not the same IP, so can't make a quantitative comparison)

Damn nice speed upgrade with no price incurred (other than time). Unless you overclock your WRT54GL (which would probably require a heatsink to be installed), I would guess that's about as much as you're going to get out of it (assuming that somebody doesn't streamline the firmwares just a tad more).

Nice gains! 😀

 

[alyarb]

Originally posted by: toonces
updated scores:

WRT54GL (w/ 1.23 Tomato)
speedtest.net = 26000/12000
torrent = 3.0Mbp/s
top peer = 1.4Mbp/s (not the same IP, so can't make a quantitative comparison)

are you confusing megabits and megabytes? speedtest.net reports its score in megabits per second, but your torrent client should use bytes per second. 3.0Mbps = 384 KB/s = not fast. This is one of those "difference between throughput and bandwidth" examples, and that is part of why we use different notation to describe each.

 

[Engineer]

Originally posted by: alyarb

Originally posted by: toonces
updated scores:

WRT54GL (w/ 1.23 Tomato)
speedtest.net = 26000/12000
torrent = 3.0Mbp/s
top peer = 1.4Mbp/s (not the same IP, so can't make a quantitative comparison)

are you confusing megabits and megabytes? speedtest.net reports its score in megabits per second, but your torrent client should use bytes per second. 3.0Mbps = 384 KB/s = not fast. This is one of those "difference between throughput and bandwidth" examples, and that is part of why we use different notation to describe each.

I'm assuming, especially with a 100Mbps connection, that the notation for torrents should be MBps instead of Mbps. I just assumed that is what toonces meant. 🙂

 

[Toonces]

Originally posted by: Engineer

Originally posted by: alyarb

Originally posted by: toonces
updated scores:

WRT54GL (w/ 1.23 Tomato)
speedtest.net = 26000/12000
torrent = 3.0Mbp/s
top peer = 1.4Mbp/s (not the same IP, so can't make a quantitative comparison)

are you confusing megabits and megabytes? speedtest.net reports its score in megabits per second, but your torrent client should use bytes per second. 3.0Mbps = 384 KB/s = not fast. This is one of those "difference between throughput and bandwidth" examples, and that is part of why we use different notation to describe each.

I'm assuming, especially with a 100Mbps connection, that the notation for torrents should be MBps instead of Mbps. I just assumed that is what toonces meant. 🙂

And you would both be correct, it's supposed to be MB/s for the torrent file of course.😛