The 127.0.0.1 problem was my windows hosts file with one dns record too much, so no wonder there was problems!
SSL transfers worked fine after I fixed that bit. I still get that network subsystem unusable in FTPRush, and in FlashFXP I get
[R] 220 -
[R] AUTH SSL
[R] 234 AUTH SSL successful.
[R] Connected. Negotiating SSL session..
[R] Connection failed (Connection lost)
[R] Delaying for 120 seconds before reconnect attempt #1
Where AUTH TLS works fine in both clients. AUTH SSL also work towards previous io versions in both clients.
Not so sure it's solely the clients fault anymore, I attempted to install latest openssl libraries, and tried flashfxp install today, in both win7, and win2k3 with the same result.. any way to look closer on what's happening when negotiating the ssl session?
I have DELAY = FALSE in ioftpd.ini, starting with empty logs
This is from clean logs:
nxmydb.log:
04-21-2010 20:04:16 ------------------------------------------------------------
<unable to log in at this point>
ioftpd.log:
04-21-2010 20:05:40 SSL: "Found certificate" "name=abcd" "Service=FTP_Service" "(Certificate_name)"
04-21-2010 20:05:40 START: "PID=1296" "CmdLine="
04-21-2010 20:05:40 PRELOAD: "begin" "../etc/default.vfs"
04-21-2010 20:06:12 LOGIN: "FTP_Service" "pion"
At that login, I'm stuck waiting for STAT -l (no disconnect, just delay) for a couple of minutes until
(20:09:01) [FTP] List Complete: 426 bytes in 84,33 seconds (0,01KB/s)
This is about 4000 folders, which each have 3 subfolders, so that leaves the total to 16000 folders. Why does it take 4 minutes, 45 seconds before login successful? I find it hard to believe that it takes that long to iterate through the subdirs on idle cpu/hdd IO..
|