io sending to "bbb" ports

[alturismo]

hi,

i have a little prob here

when the source server is ioFTPD and trying to fxp to an server
running on "bbb ports" like 22,23,80,443 the fxp transfers always
crashes after few files.

when target is also ioFTPD it looks like this
-----------------------------------------------------------------------------------
PORT x,x,x,x,0,80
200 PORT command successful.
STOR test.part05.rar
150 Opening BINARY mode data connection for test.part05.rar.
RETR test.part05.rar
150 Opening BINARY mode data connection for test.part05.rar.
426 Connection closed: Unknown error
Transfer Failed!
PASV
550 Datachannel in use, please wait and try again.
ABOR
Transfer Failed!
PASV
550 Datachannel in use, please wait and try again.
-----------------------------------------------------------------------------------

when target is rai*** it looks like that
-----------------------------------------------------------------------------------
PORT x,x,x,x,0,22
200 PORT command successful.
STOR test.part06.rar
RETR test.part06.rar
150 Opening BINARY mode data connection for test.part06.rar.
426 Connection closed: Unknown error
426 Connection closed , transfer aborted (transfer failed).(no data connection)
ABOR
Transfer Failed!
PASV
227 Entering Passive Mode (x,x,x,x,0,23)
PORT x,x,x,x,0,23
200 PORT command successful.
STOR test.part07.rar
RETR test.part07.rar
150 Opening BINARY mode data connection for test.part07.rar.
426 Connection closed: Unknown error
-----------------------------------------------------------------------------------

source server ioFTPD data port ranges tested like 1400-1500 or
40000-50000, same result

when target server is also on large port ranges its all fine.

so i tryed with another source like rai*** and target on special ports,
all working fine

so, is there a known bug when io sending to an limited data port range like described above ?

thx ahead

[ZilverZtream]

We experience the same problem when running on High Prioerety ports on our sites.

Goes fine at first.. but then you get the Unknown Error and the connection dies.

I'll look into it when I get some time.

>>zz<<

[alturismo]

thx ;)

i tryed now with 5 servers around and all the same result,

no more clues here :/

[darkone]

I'd assume that this is fixed in the next version. (I did rewrite most of both pasv & port routines :))

[darkone]

Studied this issue further, and it still exists in the latest developement version. Cause for it is not in io, but in windows socket API...

As some of you know, ioFTPD tries to bind active (PORT a,b,c,d,x,y) mode transfers to 'server listen port - 1', which is normal behaviour on unix ftp daemons (proftpd, wuftpd, ...) In other words, it tries to use same source port for every active mode transfer. Combine this with winsock, tcp lingering, fast transfers & limited amount of destination ports - and you'll end up getting errors:

Bind(1.2.3.4:5.6) + Connect(4.3.6.4:9.5) => Success
Close() => Linger
Bind(1.2.3.4:5.6) + Connect(4.3.6.4:9.6) => Success
Close() => Linger
Bind(1.2.3.4:5.6) + Connect(4.3.6.4:9.7) => Success
Close() => Linger
Bind(1.2.3.4:5.6) + Connect(4.3.6.4:9.5) => Fail, similar lingering connection exists (identical source & destination addressess)
... ~30secs ....
Lingering connections are removed by winsock.
...
Bind(1.2.3.4:5.6) + Connect(4.3.6.4:9.5) = Success
Close() => Linger

[phoenixfr]

hum ...
but this does not seem t ohappen with raidenftpd ....
what can we do against that

[alturismo]

yep, i know

only happens with ioFTP sending to high priority ports servers

all others working, tested with raiden, servu, g|ftp,... as source servers, they all work with that

lets wait prolly on a new version ;)

[alturismo]

additionel info after testing

as i understand that, what causes this problem is, as i have to use

alternative fxp upload PASV - ON in ffxp on target sites so that those high priority ports work

when this is turned on,

it only uses port command on source site while transfers, so not PASV i think

and

this only happens on fast transfers

fxp in that mode 15 meg files @90k is all fine

fxp 15 meg files @500k it crashes

fxp 50 meg files @500k it only crashes few times, most likely its working

so 50megs @500k seems to about the break even point of working/non working

as i said before, only happens when using alternative fxp upload pasv - ON on target site

source site ioFTPD normel settings and login options

may its just a little buggy when sending fast in port mode,
cause when target is "alternative ON" it looks to me source is then in port mode

ok, more tested and yes, that seems to be the problem

i tryed now only port mode on source ioFTPD (alternative FXP dl NONPASV - ON)

to several servers, and when speed is high it crashes.

no matter wich port ranges i choose on both servers, source & target

as normally it always should be PASV, but when using high priority ports,
the alternative fxp UL must be ON, and then source is prolly in port mode :/ wich makes the probs as far as i could figure out here

may that gives some more conclusions to the pros here ;)

[darkone]

Raiden does not bind active transfers to certain port - it uses random ports. If I set io to use random ports, this error does not occur.. you can emulate this behaviour by limiting local port range from you client to 1 port.

[21:52:23] LIST -la
[21:52:23] 150 Opening ASCII mode data connection for directory listing.
[21:52:23] 226 [PWD: /home/darkone/ ]-[Section: Home]-[Credits: 15.6M]-[Ratio: Unlimited]-[Speed: 220.4kb/s]
[21:52:24] PORT 192,168,1,10,4,0
[21:52:24] 200 PORT command successful.
[21:52:24] LIST -la
[21:52:24] 150 Opening ASCII mode data connection for directory listing.
[21:52:24] 426 Connection closed: Unknown error

[Unique]

You guys sure serv-u works fine?

Since serv-u also uses 1 port for active transfers...
If serv u can do it, it prolly has some routine to close the old lingering connection..

[darkone]

That wouldn't be very nice tho :) (It could result to data loss)

[alturismo]

hi,

setting random = true on source, target or both sites

didnt bring any results

same failure still

[Ren]

Im having the same problem, fxp'ng to myself with g6ftpd server on my side, at 120k :mad:

[darkone]

You don't seem to get what I'm trying to explain :p So let's try again: Error occurs when source/target ACTIVE mode is bound to certain FIXED port, and target/source PASSIVE mode port range is very low.

[stev0]

i got this problem as well.. but in flashfxp i went to the alternative fxp bit and put "download not passive" and now it works.. havnt tested on many sites but 2 out of 2 have worked when b4 it was 0 out of 2