Some tests are in Mbits/s others in Mbytes/s. Keep them the same for better understanding.
You had 4Mbits/s peak in the first post.
iwinfo reports about 94Mbits/s speed (best case scenario, not applicable though)
first iperf reported 40Mbits/s and 27Mbits/s (pretty feasible)
Before raising the distance you had ~2,55Mbytes/s =21Mbits/s and Raising the distance to 100 gives you ~3Mbytes/s= 24Mbits/sec both ways. That's an improvement.
Wired iperf gives you 111Mbytes/s=888Mbits/s (very close to theoretical speed)
Experimenting a bit with the window option and UDP can give you optimized results.
However it seems to me that your bottleneck is CPU (you can verify that by running
top when you transfer the files) and the bus between wired switch and wireless device.
As @mbo2o suggested, go ahead and upgrade the OpenWrt version, it might help things and will definitely keep you secure from vulnerabilities.
I don't think the current release will achieve higher speeds but is obviously secure and more stable
Okay sorry lleachii. I miss-read you device has to send.... I understand better now.
Guess I'll still check in the coming weeks to update OpenWRT on it, and see if it changes anything. Regardless, I'll also start to check for a new router. Guess I shouldn't ask for a recommandation on this post, as I'm sure they are plenty of other posts for that question
An example of top's output while copying a 5GB ISO file....
Mem: 35656K used, 24664K free, 752K shrd, 2416K buff, 12008K cached
CPU: 0% usr 0% sys 0% nic 58% idle 0% io 0% irq 40% sirq
Load average: 0.05 0.13 0.29 2/58 4439
PID PPID USER STAT VSZ %VSZ %CPU COMMAND
4401 3892 root R 1360 2% 1% top
1648 1 root S 7380 12% 0% /sbin/dsl_cpe_control -i05_00_04_00_4
943 1 root S 1564 3% 0% /sbin/netifd
3891 1142 root S 1216 2% 0% /usr/sbin/dropbear -F -P /var/run/dro
40 2 root SW 0 0% 0% [kworker/0:1]
3 2 root SW 0 0% 0% [ksoftirqd/0]
1288 1 root S 3224 5% 0% /usr/sbin/openvpn --syslog openvpn(Co
3117 1 root S 1644 3% 0% /usr/sbin/hostapd -P /var/run/wifi-ph
4439 943 root S 1628 3% 0% /usr/sbin/pppd nodetach ipparam wan i
1247 1 root S 1624 3% 0% /usr/sbin/uhttpd -f -h /www -r router
905 1 root S 1528 3% 0% /sbin/rpcd
1 0 root S 1404 2% 0% /sbin/procd
1704 1 root S 1360 2% 0% /usr/sbin/ntpd -n -S /usr/sbin/ntpd-h
1114 1 root S 1360 2% 0% /usr/sbin/crond -f -c /etc/crontabs -
3892 3891 root S 1360 2% 0% -ash
975 1 root S 1168 2% 0% /usr/sbin/odhcpd
1142 1 root S 1148 2% 0% /usr/sbin/dropbear -F -P /var/run/dro
896 1 root S 1040 2% 0% /sbin/logd -S 16
1309 1 nobody S 944 2% 0% /usr/sbin/dnsmasq -C /var/etc/dnsmasq
^C222 1220 nobody S 900 1% 0% /usr/sbin/tinyproxy -c /var/etc/tinyp
So... not sure what I'm supposed to see here.. I left the copy going for quite some time before checking top... All I can as the difference in top when NOT copying data is my CPU idle time which is at around 87% idle instead of 58% ...
- That's what you're looking for.
- Where is your NAS? (On LAN...on WAN...on WiFi???)
Do you ever see kworker or ksoftirqd rise?
My NAS is connected via LAN(eth) to my router, and my desktop is connected via Wifi to my router.
As for the kworker or ksoftirqd, I see them showing up once in a while while it refreshes, but %CPU isn't moving for both of them. It pretty much stays at around 0% ...
Is it possible to connect them both on the switch of the router and try to copy once again?
Also can you connect them back to back with a cable and repeat the file transfer?
I am curious if the bottleneck is on the router or on the NAS for some reason.