Re: [gentoo-user] ftp transfer dies

[Mick <michaelkintzios@gmail.com>]

2008/10/21 Robert Bridge <robert@robbieab.com>:
> On Tue, 21 Oct 2008 08:21:49 +0100
> Mick <michaelkintzios@gmail.com> wrote:
>
>> On Tuesday 21 October 2008, Paul Hartman wrote:
>> > On Tue, Oct 21, 2008 at 12:15 AM, Mick <michaelkintzios@gmail.com>
>> > wrote:
>> > > Hi All,
>> > >
>> > > Any idea why this happens:
>> > > ========================================
>> > > 150 Ok to send data.
>> > > 100% |***********************************|   224 MiB   46.74
>> > > KiB/s 00:00 ETA
>> > > 226 File receive OK.
>> > > 235279855 bytes sent in  1:21:59 (46.70 KiB/s)
>> > > local: xab remote: xab
>> > > 227 Entering Passive Mode (205,178,145,65,166,71)
>> > > 150 Ok to send data.
>> > >  34% |***********                        |   115 MiB   46.80
>> > > KiB/s 1:19:27 ETAtnftp: Writing to network: Connection reset by
>> > > peer 0% |                                   |    -1        0.00
>> > > KiB/s --:-- ETA
>> > > 500 OOPS: child died
>> > > ========================================
>> > >
>> > > It is rare that I am able to complete more than a single file
>> > > transfer before the "connection is reset by peer".  As these are
>> > > relatively large files and the upload is unattended this is
>> > > rather annoying. --
>> > > Regards,
>> > > Mick
>> >
>> > That used to happen to me when I was using a piece-of-junk D-Link
>> > router. It was one of those $29.99 consumer-grade deals. It would
>> > reboot itself constantly when it was under any kind of load. I
>> > replaced it with a $50 router with DD-WRT and things have been fine
>> > ever since. Might not have anything to do with your problem, but I
>> > figured I'd mention it. Check your router logs to see if it's having
>> > any problems.
>>
>> Thanks Paul,
>>
>> On the client side I am running a $500 professional grade router and
>> I assume that the server ISP is also running something upmarket in
>> their data center.
>>
>> On this topic the client-server arrangement straddles the Atlantic
>> ocean, so who knows how many routers and switches it jumps across.
>> That said the failure pattern is consistent:  first file always
>> transfers cleanly, then second transfer fails after a while.  Could
>> it be some configured disk/account quote, dropping transfers above a
>> certain size on the (Unix) server?
>
> Are you running through a proxy?

No, although I would love to be able to do that at work!  They only
allow port 80 to get out through the corporate gateway and probably
are running some clever filters on their Cisco routers to stop other
protocols.
-- 
Regards,
Mick