ftp transfer dies

[barchan]

Hi,

I have Wl-500gx with newest non-RC Oleg's firmware and giftd installed in /opt according to ipkg setup instructions of macsat.

Problem: ftp transfer (and probably other large files transfer too) freezes, long enough to prevent i.e. firefox to download the file. I have to use download agent to achieve any downloading at all. Problem is not visible with web pages, but may apply to them also.

I though it was a problem with my adsl modem's DHCP lease time (set originally to 60 sec) but after changing lease to 1 day problem still persists.

The ping test shows the following:



[admin@asus root]$ ping onet.pl
PING onet.pl (213.180.130.200): 56 data bytes
64 bytes from 213.180.130.200: icmp_seq=0 ttl=60 time=30.1 ms
64 bytes from 213.180.130.200: icmp_seq=1 ttl=60 time=30.7 ms
64 bytes from 213.180.130.200: icmp_seq=2 ttl=60 time=26.3 ms
64 bytes from 213.180.130.200: icmp_seq=3 ttl=60 time=30.3 ms
64 bytes from 213.180.130.200: icmp_seq=4 ttl=60 time=26.2 ms
64 bytes from 213.180.130.200: icmp_seq=5 ttl=60 time=30.2 ms
64 bytes from 213.180.130.200: icmp_seq=6 ttl=60 time=26.0 ms
64 bytes from 213.180.130.200: icmp_seq=7 ttl=60 time=30.1 ms
64 bytes from 213.180.130.200: icmp_seq=8 ttl=60 time=25.6 ms
64 bytes from 213.180.130.200: icmp_seq=9 ttl=60 time=30.0 ms
64 bytes from 213.180.130.200: icmp_seq=10 ttl=60 time=30.7 ms
64 bytes from 213.180.130.200: icmp_seq=11 ttl=60 time=29.5 ms
64 bytes from 213.180.130.200: icmp_seq=12 ttl=60 time=25.3 ms
64 bytes from 213.180.130.200: icmp_seq=13 ttl=60 time=29.3 ms
64 bytes from 213.180.130.200: icmp_seq=14 ttl=60 time=25.0 ms
64 bytes from 213.180.130.200: icmp_seq=15 ttl=60 time=29.2 ms
64 bytes from 213.180.130.200: icmp_seq=20 ttl=60 time=2017.6 ms
64 bytes from 213.180.130.200: icmp_seq=21 ttl=60 time=1018.1 ms
64 bytes from 213.180.130.200: icmp_seq=22 ttl=60 time=42.6 ms
64 bytes from 213.180.130.200: icmp_seq=23 ttl=60 time=28.3 ms
64 bytes from 213.180.130.200: icmp_seq=24 ttl=60 time=29.3 ms
64 bytes from 213.180.130.200: icmp_seq=25 ttl=60 time=28.1 ms
64 bytes from 213.180.130.200: icmp_seq=26 ttl=60 time=29.2 ms
64 bytes from 213.180.130.200: icmp_seq=27 ttl=60 time=28.0 ms
64 bytes from 213.180.130.200: icmp_seq=28 ttl=60 time=29.0 ms
64 bytes from 213.180.130.200: icmp_seq=29 ttl=60 time=27.6 ms
64 bytes from 213.180.130.200: icmp_seq=30 ttl=60 time=28.6 ms
64 bytes from 213.180.130.200: icmp_seq=31 ttl=60 time=24.7 ms
64 bytes from 213.180.130.200: icmp_seq=32 ttl=60 time=28.5 ms
64 bytes from 213.180.130.200: icmp_seq=79 ttl=60 time=3017.3 ms
64 bytes from 213.180.130.200: icmp_seq=80 ttl=60 time=2017.7 ms
64 bytes from 213.180.130.200: icmp_seq=81 ttl=60 time=1018.1 ms
64 bytes from 213.180.130.200: icmp_seq=82 ttl=60 time=40.3 ms
64 bytes from 213.180.130.200: icmp_seq=83 ttl=60 time=25.8 ms
64 bytes from 213.180.130.200: icmp_seq=84 ttl=60 time=29.9 ms
64 bytes from 213.180.130.200: icmp_seq=85 ttl=60 time=25.7 ms
64 bytes from 213.180.130.200: icmp_seq=86 ttl=60 time=29.6 ms
64 bytes from 213.180.130.200: icmp_seq=87 ttl=60 time=25.5 ms
64 bytes from 213.180.130.200: icmp_seq=88 ttl=60 time=29.5 ms
64 bytes from 213.180.130.200: icmp_seq=89 ttl=60 time=25.1 ms
64 bytes from 213.180.130.200: icmp_seq=90 ttl=60 time=29.2 ms
64 bytes from 213.180.130.200: icmp_seq=91 ttl=60 time=30.2 ms
64 bytes from 213.180.130.200: icmp_seq=92 ttl=60 time=29.0 ms

--- onet.pl ping statistics ---
130 packets transmitted, 43 packets received, 66% packet loss
round-trip min/avg/max = 24.7/236.9/3017.3 ms

As you may see, a dozen or so pings are normal, and then there is a freeze. Then back to normal, freeze etc. so it goes.

Do you have any idea what's happening?

Barchan

[Styno]

And what about the DHCP lease time of your asus router?

Perhaps something is putting a strain on your usb port? Like a webcam or virsusscanner scanning files on your usb storage?

I have seen lots of lost connections doing ftp transfers from the usb storage while copying large files over SMB to the attached usb storage. This might point to a scheduling problem somewhere....

[barchan]

Thanks for replying.

There is a harddrive connected to one USB port, but it was idle during FTP transfers.

Actually, I think FTP transfers were affected by the connection problem which ping test made visible, i.e. a lot of packets are being lost for some strange reason.

giFT was not an issue, the problem remains after switching it off.

The client lease time for LAN is 86400 sec i.e. the same as for WAN.

Very strange problem. Never had any similar issue with DSL line, so it is not the reason. I also believe DSL modem (Dlink 302T) is all right, when connected directly to computer I did not observe any freezes.

I will be grateful for further assistance :-)

Barchan