openssh unix dev - Jan 2012 - scp should not report 100% until the last byte is transferred

Hello,

just my opinion, but yesterday I run into a case, where after 7 hours of 
transferring a file (a little bit greater than 1 GB) the scp reported 100% and 
after that exited with an error (AIX 7.1):

$ tsocks scp foo user at host:/tmp
...
100% 1101MB  44.2KB/s 448.0KB/s 7:05:38    
scp: /tmp/foo: File too large

Probably it should not be rounded up to 100% ?

-- 
MfG/Sincerely
Toralf F?rster
pgp finger print: 7B1A 07F4 EC82 0F90 D4C2 8936 872A E508 7DB6 9DA3

Toralf F?rster wrote:
> just my opinion, but yesterday I run into a case, where after 7 hours of 
> transferring a file (a little bit greater than 1 GB) the scp reported 100%
and
> after that exited with an error (AIX 7.1):
> 
> $ tsocks scp foo user at host:/tmp
> ...
> 100% 1101MB  44.2KB/s 448.0KB/s 7:05:38    
> scp: /tmp/foo: File too large
> 
> Probably it should not be rounded up to 100% ?
The scp protocol is very primitive, and I'm not sure that an error
can be detected at all until after 100% has been sent.

Recommend use SFTP.


//Peter

On Thursday, 12. Jan 2012, Toralf F?rster wrote:
> just my opinion, but yesterday I run into a case, where after 7 hours of
> transferring a file (a little bit greater than 1 GB) the scp reported
> 100% and after that exited with an error (AIX 7.1):
> 
> $ tsocks scp foo user at host:/tmp
> ...
> 100% 1101MB  44.2KB/s 448.0KB/s 7:05:38
> scp: /tmp/foo: File too large
> 
> Probably it should not be rounded up to 100% ?
Perhaps another part of the problem is that it doesn't wait for byte 
position acknowledgments from the remote side ...

I've seen it fill up the TCP buffers, and then stay at 100% for ~30 seconds 
until the buffers were drained over the slow connection.


Regards,

Phil