scp "lost connection" but ssh works fine

A server I can ssh into fine has begun refusing to scp.

$ scp ~/tmp/foo :~/tmp/ lost connection 

With scp -v -v I can see the connection succeeds and the transfer appears to succeed, but no file appears on the other side.

OpenSSH_5.9p1, OpenSSL 0.9.8r 8 Feb 2011 debug1: Reading configuration data /Users/schwern/.ssh/config debug1: /Users/schwern/.ssh/config line 1: Applying options for * debug1: Reading configuration data /etc/ssh_config debug1: /etc/ssh_config line 20: Applying options for * debug2: ssh_connect: needpriv 0 debug1: Connecting to testcurrent01.dev.liquidweb.com [10.30.152.254] port 22. debug1: Connection established. debug1: identity file /Users/schwern/.ssh/id_rsa type -1 debug1: identity file /Users/schwern/.ssh/id_rsa-cert type -1 debug1: identity file /Users/schwern/.ssh/id_dsa type -1 debug1: identity file /Users/schwern/.ssh/id_dsa-cert type -1 debug1: Remote protocol version 2.0, remote software version OpenSSH_4.3 debug1: match: OpenSSH_4.3 pat OpenSSH_4* debug1: Enabling compatibility mode for protocol 2.0 debug1: Local version string SSH-2.0-OpenSSH_5.9 debug2: fd 3 setting O_NONBLOCK debug1: SSH2_MSG_KEXINIT sent debug1: SSH2_MSG_KEXINIT received ...lots of authentication details... debug1: Enabling compression at level 6. debug1: Authentication succeeded (publickey). Authenticated to ([1.2.3.4]:22). debug2: fd 5 setting O_NONBLOCK debug2: fd 6 setting O_NONBLOCK debug1: channel 0: new [client-session] debug2: channel 0: send open debug1: Entering interactive session. debug2: callback start debug2: client_session2_setup: id 0 debug2: fd 3 setting TCP_NODELAY debug1: Sending environment. debug1: Sending env LANG = en_US.UTF-8 debug2: channel 0: request env confirm 0 debug1: Sending command: scp -v -t -- ~/tmp/ debug2: channel 0: request exec confirm 1 debug2: callback done debug2: channel 0: open confirm rwindow 0 rmax 32768 debug2: channel 0: rcvd adjust 2097152 debug2: channel_input_status_confirm: type 99 id 0 debug2: exec request accepted on channel 0 debug2: channel 0: rcvd eof debug2: channel 0: output open -> drain debug2: channel 0: obuf empty debug2: channel 0: close_write debug2: channel 0: output drain -> closed debug1: client_input_channel_req: channel 0 rtype exit-status reply 0 debug2: channel 0: rcvd close debug2: channel 0: close_read debug2: channel 0: input open -> closed debug2: channel 0: almost dead debug2: channel 0: gc: notify user debug2: channel 0: gc: user detached debug2: channel 0: send close debug2: channel 0: is dead debug2: channel 0: garbage collecting debug1: channel 0: free: client-session, nchannels 1 debug1: fd 0 clearing O_NONBLOCK debug1: fd 1 clearing O_NONBLOCK Transferred: sent 4576, received 2520 bytes, in 0.0 seconds Bytes per second: sent 167737.0, received 92372.6 debug1: Exit status 0 debug1: compress outgoing: raw data 135, compressed 121, factor 0.90 debug1: compress incoming: raw data 66, compressed 52, factor 0.79 lost connection 

It is a CentOS 5.9 machine.

Things I've checked...

• I have permission to write to that directory.
• The user has a sensible shell (/bin/bash).
• I tried moving my ~/.ssh/config out of the way.
• scp'ing to that machine from others with entirely different operating systems also fail.
• The disk is not full.
• Restarting sshd.

/var/log/secure contains...

Apr 4 14:23:22 some sshd[12576]: Postponed publickey for user from 1.2.3.4 port 33581 ssh2 Apr 4 14:23:22 some sshd[12575]: Accepted publickey for user from 1.2.3.4 port 33581 ssh2 Apr 4 14:23:22 some sshd[12575]: pam_unix(sshd:session): session opened for user user by (uid=0) Apr 4 14:23:22 some sshd[12575]: pam_unix(sshd:session): session closed for user user 

What might I check next?

4 Answers

Had the same issue.

If you did a minimal install of Centos, it only install the openssh and openssh-server packages but not the openssh-clients. sudo yum install openssh-clients will fix your issue.

scp works by making an ssh connection to the remote host, then launching another copy of the scp program on that host. The two scp instances communicate through the ssh connection to perform the file transfer.

"lost connection" is printed by the local scp program when the ssh connection drops prematurely. The usual reason for that is the scp program on the remote host either failed to start or else it exited prematurely. This could have happened because the scp program doesn't exist on the remote host, or it's not in your command PATH, or it's not marked executable, or it crashed after starting, or something along those lines.

If you happen to be running the ufw package and have port 22 set with sudo ufw limit 22, you will need to open it up with:

sudo ufw allow 22 

Check status with:

sudo ufw status 

We recently had this issue on one of our systems.

We could appropriately ssh onto the host server, but discovered we could not ssh from the server back onto the machine. This is a good place to investigate, if you cannot do this then you will not be able to use SCP.

In our case, somehow (perhaps a botched install) had replaced our ssh binary files with 0 byte empty files. Whenever "ssh" was executed, nothing happened.

By reinstalling openssh-clients, we rectified the binaries and scp started to work.

yum reinstall openssh-clients