Kind of a weird issue here.
We have a lot of data dropped on our SFTP, and sometimes we get a lot of files written with 0 bytes. Upon checking the remote users SFTP Client (FileZilla) it shows the transfer was successful for the same amount of bytes, but when looking at SFTP logs it shows that the SFTP is opened with WRITE flags, but nothing happens until eventually the session is broken off about 2 hours later (I assume when they close FileZilla). Here is a excerpt from the logs, some names redacted.
Mar 25 07:02:12 SERVER internal-sftp: realpath "/SERVER/IN/20150324/Part_4/20150324_Upload Tonight_SF_Table_4/20150321_SF_TT_04_KEVIN_0225.jpg" Mar 25 07:02:13 SERVER internal-sftp: open "/SERVER/IN/20150324/Part_4/20150324_Upload Tonight_SF_Table_4/20150321_SF_TT_04_KEVIN_0225.jpg" flags WRITE,CREATE,TRUNCATE mode 0666 Mar 25 09:13:40 SERVER internal-sftp: forced close "/SERVER/IN/20150324/Part_4/20150324_Upload Tonight_SF_Table_4/20150321_SF_TT_04_KEVIN_0225.jpg" bytes read 0 written 0
But when looking at the FileZilla transfer window it shows up as a successful transfer and the file sizes matches the real file size.
A screenshot of this:
As far as I know SFTP is supposed to have built in error-checking. But this does not appear to be the case. Does anyone know why this could be occurring? It happened on quite a few files today.