Tag Info

New answers tagged

0

I have recently experienced this with Gerrit's SSH interface. The problem was that my local SSH agent offered up a bunch of different keys to the Gerrit server, and after some limit the server just refused to accept further keys (but still replied with the Server accepts key). I don't know if this behavior is specific to Gerrit or a generic OpenSSH thing. ...


0

It seems the SSH key that you used to login into the board is missing. The following command lists the set of keys installed on your hosts. ssh-add -l If you have console access to the board then you can edit the sshd.conf file in the board to disable password access. If not, get in touch with the admin to install the respective SSH key in your host. Try ...


0

You do need to use SOCKS5. This article has details for PuTTY. You tried to set up tunneling by running a command on your server; instead, you need to configure PuTTY. Before connecting, in PuTTY, under Connection > SSH > Tunnels, you need to add a dynamic tunnel, e.g. on port 8080, then configure Firefox.


0

This has been resolved. There were TWO access lists - the first that was commonly known about (ACL) which lists the servers that a user can connect to, and the second previously unknown list of servers that are only allowed access to certain users, ie the reverse of an ACL. With the servers removed from this restrictions list, access was allowed for all ...


0

The free version doesn't support SSH so you should use the instructions provided on the NoMachine website to know how to tunnel NX over an SSH connection.


0

tl;dr: It was a DNS leak. Thanks to @user2675345 for the tip off! Here is the page I got when I accessed a blocked site: I first tried pinging a few websites and saw that their IP addresses were the same. As can be seen in this image: both youtube.com and metacafe.com, which are both blocked, have 176.12.107.179 as their IP. Unsurprisingly, navigating ...


0

You've probably set it as an HTTP proxy, not a SOCKS proxy. Make sure you've selected it as a SOCKS proxy.


2

Try SFTP. It's the protocol used by SSH to transfer files and is supported by VLC in the Open Network Stream menu by typing sftp://<server>/<path to file> and then authenticating.


1

Check user A's .ssh/config file. He may have a Host entry in that file for the hostname at issue, which is causing ssh to try a different port or IP address for that hostname. For example, if .ssh/config had these lines: Host a.example.com Hostname b.example.com Port 42 Then running "ssh a.example.com" would try try to connect to b.example.com ...


0

Set the Socks proxy only, leave the others blank (do not check the box for use the same..)


0

In the end, user@domain.local@host actually worked fine.


0

Finally I solved this problem The problem is I set the Chrome proxy setting wrong. I write the proxy host and port in the "HTTP proxy" line. This is wrong. I should write them just in the "SOCKS proxy" line and left the other blank


0

This command worked on my system: virt-manager ; pkttyagent --process $(pgrep virt-manager)


0

You may be running ssh-keygen on the wrong file. ssh-keygen -y operates on a private key file. ".pub" files normally contain the public key. You probably have a file there named my_key, without any extension, and it ought to be mode 0700. That is the file which should contain the private key. To directly answer your question, SSH keys are normally used to ...


0

scp -i <path-to-public-key> <remote-user@remote machine>:<remote-path-to-file> <local-destination> For example scp -i Downloads/xyz.pem ubuntu@xyz.com:/home/ubuntu/sample.csv Desktop/


1

There is another possible approach. You can use placeholders in the IdentityFile option in ssh_config. As per man ssh_config: %d—local user's home directory %u—local user name %l—local host name %h—remote host name %r—remote user name I use it like this (in the global ssh_config file): IdentityFile ~/.ssh/%r@%h This means my private key files are ...


1

If you want to offer the same keys to all hosts, load them into a SSH agent using ssh-add. Many Linux distributions start one automatically – try ssh-add -l to check if it's running, then load your keys: ssh-add ~/.ssh/id_rsa ~/.ssh/surnameg etc. If the agent not started automatically, put the following in your ~/.profile: agent_running() { [ ...


0

0644 in not supposed to be too open for a public key, but is too open for your private key. Your private key should have permission 0600 while your public key have permission 0644. By the way, you should also take care of the permission on .ssh folder. It should has the permission 0700, so that only you, the owner, has control over the folder. As to your ...


0

Here's my solution: function! WinCpy () let s:wincpy = exists('s:wincpy') ? !s:wincpy : 1 if !s:wincpy :q else silent exec "normal \<C-w>S" silent exec "normal \<C-w>T" silent exec "set nonu" silent exec "set wrap" endif endfunction vnoremap <leader>y :call WinCpy()<CR> ...


1

you might have better luck if you just used phpseclib, a pure PHP SSH client, to emulate libssh2: https://github.com/phpseclib/libssh2-compatibility-layer


0

Just like @user2675345 is saying, you should probably check the DNS proxy settings if your browser has any. Follow these steps to enable DNS lookups through a proxy in Firefox: enter about:config in the address bar search for proxy set network.proxy.socks_remote_dns to true I used to be a Chrome user myself, but switched to Firefox when I noticed that ...


0

I love git, but unfortunately, git is not right tool for this task. Git was designed to very efficiently keep change history for mostly text content repositories. While git does support keeping binaries, it will have to keep them forever in history so you can checkout to any revision, which is very expensive in terms of disk space. Also, assuming that your ...


0

I realise that this is not really answering your question, but... wouldn't rsync be far easier to keep two folders in sync?


1

You can check with ping if your machine comes up and then fire off one or several ssh commands. Note that you will need to setup certificate-based authentication on your host. Here's a little script based on GNU/Linux' BASH but I guess you can convert it to whatever Mac OS X is using. #/bin/bash hostup=0 while [ $hostup -ne 1 ]; do ping -c 4 ...


0

Well on Linux it's pretty straightforward, first you run e.g.: xhost +localhost On the remote machine, and then from a remote SSH session you can just set DISPLAY, e.g.: DISPLAY=:0.0 program and its options I don't know if Windows would behave the same way, it doesn't run X so there would certainly be some other trickery involved behind the scenes. Is ...


0

you need to configure your host machine to run ssh. Windows doesnt, by default, has a ssh service running. You need to set-up and run a ssh daemon service. Follow this article: http://docs.oracle.com/cd/E24628_01/install.121/e22624/preinstall_req_cygwin_ssh.htm


0

Here are advices to secure your ssh connection: If you don't need to access your ssh server throw internet, allow only connections from your LAN or even better from specific IP If you need to access your ssh server throw internet, change your ssh port Don't use the root account to connect to your ssh server Use a private key to connect to your ssh server ...


1

I'd suggest another tactic - While there are tools that allow for load balancing, it may not work for your use case as it is. I suggest having a session thats resilient to disconnections Firstly, I'd run my commands in screen or tmux - that way your commands keep running even if you get disconnected. You can then log back in and re-attach the session. You ...


0

I had the same issue. MINGW's SSH is actually home for the path /home//.ssh Put your SSH configs there and it'll work


0

Check your SSH client config. TCPKeepAlive should not be set to no for the client.


2

Putty is a good place to start. Alternatively (or in addition to), WinSCP will help you maintain that FTP/Explorer view you're probably used to, and if you're not trying to learn command lines. WinSCP will requires Windows, and you can use this to connect to any SSH server.


1

For uploading files you could use Filezilla as a client and use it like any other FTP programme. To set it up, first click on the Site Manager button (1) and enter the details provided by your administrator. Make sure to select SFTP as the protocol (3) and 22 as the port (4). Change Logon Type to Normal (5) and enter your username (6) and password (7). ...


0

Pwnat seems to be unauthenticated which I'd consider a major security risk. If you control at least one of the two NAT/Firewalls just setup port forwarding/translation for a much more secure setup.


0

If you want to add your public key to authorized_keys on a remote machine, an answer may be ssh-copy-id (e.g. part of openssh-clients).


0

Are you sure your local Mac is reachable at that IP address, from the remote machine? For example, if you had an RFC 1918 private address such as 192.168.x.y on your local Mac because it was behind a NAT gateway, and the remote machine was on the public side of the NAT gateway, then the remote machine wouldn't be able to reach your local Mac at that address, ...


3

Do I understand correctly that you're using SSH to connect to the remote host, then you're running scp on the remote host to copy the file back to the local host? Without knowing anything about these two hosts, it's perfectly possible that your local host can make SSH connections to the remote, but the remote can't make SSH connections to the local. If you ...


1

One option to make it work with a single SSH session is to use a variable to store your bash file in rather than copy it. $ mybash=`cat mybash` $ ssh -t 127.0.0.1 "echo $mybash > /tmp/mybash; bash --rcfile /tmp/mybash ; rm /tmp/mybash" Seems to work for me.


0

Have you tried to ssh in manually? ssh -i pubkey root@192.168.200.122 My only guess is for whatever reason the client's known_hosts got changed, deleted, etc. Or the entry in it doesn't match anymore which is possible. If openssl got changed (upgraded?) on the server side, then the known_hosts entry won't match. And so when you ssh, even with a ...


1

Whenever I have seen ssh delays like the above, it was tied to DNS. 1) Check to see if "UseDNS No" is in your /etc/ssh/sshd_config file 2) if it is not (default is Yes and will be commented), backup your /etc/ssh/sshd_config file 3) Add UseDNS No 4) restart sshd (do not close your current ssh terminal if that is how you logged in) 5) open new putty, see ...


1

We use PUTTY on windows to SSH into some 300 linux servers across a diverse network environment. When we have a server that is lagging on the SSH login but then behaves normally once the session is established, we have resolved the issue as follows. -- Put an entry in the /etc/hosts file of the VM for the VM itself so that the SSH server on the linux guest ...


0

Sounds to me like your ISP blocks port 22, maybe due to Heartbleed, most likely to discourage free unix shell providers. I would suggest mapping to a different external port, see if that makes a difference.


21

See this manpage for time.conf user1 ; * ; Wd0000-2400 | Wk1800-0800 would allow the user user1 evenings on weekdays (wk), and all days on weekends (wd), and deny them the rest of the time. yours would be something like * ; * ; Al0800-0900


0

What I did with TortoiseSVN is: try disconnecting your network, you will be asked to take tortoise offline, choose offline mode do a cleanup connect your network and try updating... That solved a same problem for me...


1

ssh theotherbox -l user@theotherbox This isn't the right syntax to specify a username and a hostname. ssh may be using "user@theotherbox" as a username, which is likely to fail. The correct syntax is either: ssh -l user host or ssh user@host


1

As Ignatio suggested this can be done with grep -v. Here is a example which removes the key containing some unique string or just deletes the authorized_keys file when no other key remains. if test -f $HOME/.ssh/authorized_keys; then if grep -v "some unique string" $HOME/.ssh/authorized_keys > $HOME/.ssh/tmp; then cat $HOME/.ssh/tmp > ...


0

Change you made in sshd_config shouldn't have influenced TeamPass. Problems you are seeing are result of something else, unless you are using SSH tunnel to connect to TeamServer. Anyway, one thing you can try is to enable following setting in your sshd_config: AllowAgentForwarding yes AllowTcpForwarding yes X11Forwarding yes and see if it helps. Don't ...


4

The screencap you've posted shows an "end port" of 23. It's possible that this forwards a range of ports, but it's equally possible that start and end port refer to "outside port" and "inside port" respectively, in which case, you've got a problem. Try changing the "end port" to 22. Also, check your /etc/hosts.{allow,deny}.


1

Try using ssh -v (or -vvv) and watch the order the ssh keys are tried. If id_rsa comes first something went wrong with your ssh_config. Try using ssh -i /path/to/id_lynx If this works check your ssh_config for IdentityFile <- should be a fully specified path, IdentitiesOnly yes man 5 ssh_config says: IdentitiesOnly Specifies that ssh(1) ...


0

My key was not being forwarded, turned out I had started the SSH agent in a different terminal window, so the $SSH_AUTH_SOCK environment variable was not available in the terminal in which I was making the connection. So if you are starting the agent manually, make sure you make the connection in the same terminal session.


0

It is a text file. You can easily edit with vi(m) and simply delete the line in question (dd), and save the file (wq). But if there is a specific command to remove a host, that's probably the safest method.



Top 50 recent answers are included