router veroorzaakt langzaam inet

Pagina: 1
Acties:

  • indexample
  • Registratie: April 2002
  • Niet online
ik heb voor nederlandse begrippen (en zelfs voor belgische begrippen) redelijk slecht internet (kotnet). Toen we vorige maand werden aangesloten, stond ik dan ook niet te kijken dat sommige pagina's traag laadden, of plaatjes het niet deden etc. Met name dingen als plaatjes van imageshack, hotmail, op sommige forums laad ie maar de helft (als ik een reply typ, kan ik niet op send drukken want die knop ontbreekt...???), ranglijsten op Clanbase laad ie (soms) niet, dus wel de pagina maar zonder de tabel.Ook Teamspeak hapert als een gek, ik versta mensen amper en zij mij ook niet.

Echter, vandaag kwam ik erachter dat als ik de router (linksys wrt54g) er tussenuit haal, alles het perfect doet! Welke instellingen zouden dit kunnen vertragen?

Deze problemen heb ik zowel over wire als wireless.

geprobeerd:

dhcp uitschakelen (doet internet het in zn geheel niet meer, geen verbinding).

een ping naar google.nl geeft dit:
PING 66.249.93.104 ( 66.249.93.104 ) : 56 data bytes
64 bytes from 66.249.93.104: icmp_seq=0, ttl=245 times=20. ms
64 bytes from 66.249.93.104: icmp_seq=3, ttl=245 times=20. ms
64 bytes from 66.249.93.104: icmp_seq=4, ttl=245 times=30. ms
--- 66.249.93.104 ping statistics ---
packets transmitted = 5 , packets received = 3 packet loss = 40%
round-trip min/avg/max = 20/23/30

een traceroute naar google.nl geeft dit:
1 traceroute to 66.249.93.104 (66.249.93.104) ,30 hops max,40 byte packet
2 10.91.135.254 (10.91.135.254) 10. 0 ms 0. 0 ms 0. 0 ms
3 192.168.25.4 (192.168.25.4) 0. 0 ms 30. 0 ms 0. 0 ms
4 192.168.25.130 (192.168.25.130) 10. 0 ms 0. 0 ms 10. 0 ms
5 134.58.253.254 (134.58.253.254) 0. 0 ms 10. 0 ms 0. 0 ms
6 193.191.9.9 (193.191.9.9) 0. 0 ms 20. 0 ms 10. 0 ms
7 193.191.1.173 (193.191.1.173) 0. 0 ms 10. 0 ms 0. 0 ms
8 193.191.1.2 (193.191.1.2) 0. 0 ms 30. 0 ms 10. 0 ms


hopelijk heb ik voldoende info gegeven en ligt er een oplossing in het verschiet :)

zo niet zal ik meer info geven.

[ Voor 3% gewijzigd door indexample op 23-11-2006 16:39 ]


  • Kosh66
  • Registratie: Oktober 2002
  • Laatst online: 08-02 22:49
Packet loss duidt meestal op een defect in de hardware....
Check je kabels en als dat niet baat, ruil je router om.

  • indexample
  • Registratie: April 2002
  • Niet online
ook met een firmware upgrade voor de router heb ik packet loss


hmm, heb hier iets gevonden over packet loss op de wrt54g v5
For those whose got packet losses after running the WRT54G (every version)for a few hours....

I got the same problem with my WRT54G v1 and thought my router was end of life. So I bought a new one, unfortunately its a WRT54G v5 with VxWorks running. This v5 has even more packet losses. So I brought back the v5 to the store and got my money back. Bought another router from USRobotics and got the same problem. After that I finally relize its not the Linksys's problem. But my own.

The problem is the default settings on the firmware is to clear the buffer every 5 days. But if you run some bittorrent, you make a lots of connections and the router's buffer will be overflowed after af few hours running some bittorrent files.

My solution was to flash it with a third party firmware, in my case the HyperWRT 2.1b1, that firmware got an option to edit the startup and firewall script in the webinterface. Then I went to search for more information and found a script to put on the start up that makes the router cleares the buffer every 30 minutes (or you can set it to 10 minutes if you like).

Now the router is running very stable, uptime almost 30 days and running non-stop 10 or more torrents at the same time.

The HyperWRT firmware can be found on www.hyperwrt.org and the startup script you should put this in...

code:
1
2
3
4
5
6
7
echo 4096 > /proc/sys/net/ipv4/ip_conntrack_max
echo 1 > /proc/sys/net/ipv4/icmp_echo_ignore_broadcasts
echo 1 > /proc/sys/net/ipv4/icmp_ignore_bogus_error_responses
echo 512 > /proc/sys/net/ipv4/neigh/default/gc_thresh1
echo 2048 > /proc/sys/net/ipv4/neigh/default/gc_thresh2
echo 4096 > /proc/sys/net/ipv4/neigh/default/gc_thresh3
echo "600 1800 120 60 120 120 10 60 30 120" > /proc/sys/net/ipv4/ip_conntrack_tcp_timeouts


that 1800 in the last code is the number of seconds that it will clear the buffer. 1800 means 30 minutes. You can set it to anything you wish, at my case 1800 is working fine.

[ Voor 4% gewijzigd door indexample op 23-11-2006 17:06 ]