Connection lost past ing data i nto an ssh session v ia a cygwi n window t o NetApp‏

[Bill Shaffer <billshaf at hotmail dot com>]

[Resending with a subject. If the last one made it to the list, I apologize profusely.]

Okay. This one is pretty esoteric; I apologize.

I am having problems pasting data (normal cut-and-paste operations) via an ssh connection if the source in a cygwin terminal or xterm, and the _final_ destination is a NetApp controller. I say final destination because if I ssh to an intermediate host, then ssh to the NetApp, I see the same problem.

The problem I see is that, while logged into the NetApp, if I paste any data (I've seen it with as few as 4 characters), I will see some of the characters (a varying number) in the window at the NetApp prompt, then will get a "Connection to host closed", and the rest of the characters show at the command line where the ssh session originated. The NetApp logs this error:
[telnet.socket.timeout:warning]: Telnet daemon socket timed out for the SShv2 connection.
and occasionally this one:
[openssh.socket.shutdown:info]: sshd_shell: socket shutdown() failed: Invalid argument.]

I see this problem if I type really fast into the ssh window as well. It's almost like I'm over-running a buffer or something.

I run a packet trace, and see that a RST is being sent by the NetApp.

If I use putty ssh instead of cygwin, I have no problem. Other users, who do not use Cygwin, do not see a problem. If I ssh to that intermediate host and telnet into the NetApp, I see no problem.

I did NOT see this problem when the NetApp was running ONTAP 8.0. As soon as I upgraded to ONTAP 8.1, I started seeing this, across multiple hosts, hardware types, and physical locations/networks.

My current version of Cygwin is 1.7.9. I upgraded all the current Cygwin packages on a different computer and had the same issues (I don't have that computer with me or I would give that version; I upgraded within the last few weeks, so it should still be current).

My assumption (still) is that something in ONTAP changed, and that it is a NetApp issue, since it's seems clear to me that it is NetApp that is terminating the connection. NetApp support says they've exhausted options on their side, insist that they are compliant with the interface standards, and urge me to use a different client.

I'm hoping that someone will either recognize this behavior, or give me some other troubleshooting options, or recommend something to go back to NetApp with.

I've attached a slightly edited cygcheck.out report. I'd be happy to provide any information that I've missed here.

Thanks for your time.
Bill

Attachment: cygcheck.out
Description: Binary data

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple