I installed the latest Google Cloud Deep Learning VM Image today, after VM was launched, I was able to do sudo -i successfully via SSH web.
Once I login, I start my Tensorflow model training running in background (Using &). Few hours later I'm unable to login as root.
I get the following message:
We trust you have received the usual lecture from the local System
Administrator. It usually boils down to these three things:
#1) Respect the privacy of others.
#2) Think before you type.
#3) With great power comes great responsibility.
[sudo] password for my_username:
I tried:
sudo -i
su sudo -i
su root
I was able to replicate the issue. Any suggestions?
This issue was caused due to an internal Google side and removes the user from “Google-sudoers” group. For all affected instances, I suggest following the below workaround until the permanent fix has been rolled out.
Use a different username:
If using browser SSH window, click on the settings icon (top right), and click change Linux name in the drop down.
Using the SDK
$ gcloud compute ssh newusername#instance
Enable OS Login on the instance (set "enable-oslogin=True" in metadata) and per this article
You can track the permanent fix by following the Public Issue tracker.
The original answer:
Maybe the solution will be to add a SSH Key for Google Cloud Console and log in with another SSH client.
Additional answer:
I do not know why, but sometime the user suddenly stopped being a member of the google-sudoers group...
Then it's enough add your user to this group by some other user with administrator privileges to this group:
# usermod -G google-sudoers your_user_name
of course, if there is such a user...
I'm just starting to work with AWS and I'm working on getting Anaconda set up on my AWS EC2 Instance. The entire set up process worked well but i'm getting stuck with inputting my password in jupyter.
FYI, here is the tutorial I have been using: https://chrisalbon.com/jupyter/run_project_jupyter_on_amazon_ec2.html
I manually changed the password in the ipython config file:
Even when inputting testing123 I get an invalid error message.
Have you guys dealt with this? Any help would be much appreciated.
I think you have understood the whole process but got confused in password section so I am only gonna point out what you did wrong.
Assuming You followed below blog as you said:
https://chrisalbon.com/jupyter/run_project_jupyter_on_amazon_ec2.html
In the section
Create a password for jupyter notebook
1 ipython
2 from IPython.lib import passwd
3 passwd()
4 'sha1:98ff0e580111:12798c72623a6eecd54b51c006b1050f0ac1a62d'
5 exit
I have added the line numbers for the sake of better explanation:
when you write passwd() which is in 3rd line
The AWS expects you to input a password and again ask you to verify the password you have given and then it gives you a SHA version of your password. like on 4th line: 'sha1:98ff0e580111:12798c72623a6eecd54b51c006b1050f0ac1a62d'
Now you have to copy this and keep it with yourself to paste it later in ipython config file.
so you should use below lines in your config to make it work for you:
c = get_config()
# Kernel config
c.IPKernelApp.pylab = 'inline' # if you want plotting support always in your notebook
# Notebook config
c.NotebookApp.certfile = u'/home/ubuntu/certs/mycert.pem' #location of your certificate file
c.NotebookApp.ip = '*'
c.NotebookApp.open_browser = False #so that the ipython notebook does not opens up a browser by default
c.NotebookApp.password = u'sha1:98ff0e580111:12798c72623a6eecd54b51c006b1050f0ac1a62d' #the encrypted password we generated above
# It is a good idea to put it on a known, fixed port
c.NotebookApp.port = 8888
im using EMR and wanted to use jupyter(ipython) so i added to the cluster the bootstrap action:
s3://elasticmapreduce.bootstrapactions/ipython-notebook/install-ipython-notebook
I performed the port tunelling to access jupyter from my local host and works fine, but it is asking for a login password, tried empty, tried hadoop, but no luck, does any body knows what is the jypyter password?
I ran into this problem as well when I used the same bootstrap action. I tried adding in Args=[--password, jupyter] which I also could not get working. That was from this aws forum:
Name='Install Jupyter notebook',Path="s3://aws-bigdata-blog/artifacts/aws-blog-emr-jupyter/install-jupyter-emr5.sh",Args=[--r,--julia,--toree,--torch,--ruby,--ds-packages,--ml-packages,--python-packages,'ggplot nilearn',--port,8880,--password,jupyter,--jupyterhub,--jupyterhub-port,8001,--cached-install,--notebook-dir,s3://<your-s3-bucket>/notebooks/,--copy-samples]
What I did instead was to follow these instructions for installing anaconda directly in the EMR instance using the CLI. If you follow the first part you should be able to get it up and running. To summarize here:
ssh into your master emr instance using the .pem file you saved
once there's you'll want to install anaconda using super user priveledges: sudo wget http://repo.continuum.io/archive/Anaconda3-4.1.1-Linux-x86_64.sh. Then bash Anaconda3–4.1.1-Linux-x86_64.sh
Make sure you're using the anaconda version of python: which python
If you're not, specify your source: source .bashrc
Now make a jupyter config file: jupyter notebook --generate-config
cd into the jupyter folder: cd ~/.jupyter/
update the config file: vi jupyter_notebook_config.py
In the config file add the following lines:
c = get_config()
c.NotebookApp.ip = '*'
c.NotebookApp.open_browser = False
c.NotebookApp.port = 6789 <---pick whichever port you want
exit out of the config editor and run jupyter via: jupyter notebook
this should run a notebook with no active kernels (for now). But it will give you the token you're looking for: http://localhost:6789/?token=xxxxxx
Leave this running, and open a new terminal window. Now you'll want to tunnel to the EMR instance per this aws blog post (make the port the same as the one you specified in the config file). ssh -o ServerAliveInterval=10 -i <<credentials.pem>> -N -L 8192:<<master-public-dns-name>>:8192 hadoop#<<master-public-dns-name>>
Opening localhost:6789 in the browser should prompt you with the jupyter page to enter your password or token. Enter the token that was generated in the above step and you should be good to go.
Hope this helps! There might be a less convoluted way, but this is what ended up working for me.
I am trying to connect a remote machine in python. I used telnetlib module and could connect to machine after entering login id and password as
tn = Telnet("HOST IP")
tn.write("UID")
tn.write("PWD")
After entering password, the terminal connects to the remote machine which is a linux based software [having its own IP address(HOST IP).]
Then after If I try to give a command e.g. tn.write("cd //tmp/media/..) to go to its various folders then it does not work and when checked to see what the screen is showing with
tn.read_very_eager()
error comes up as :
""\r\n\r\n\r\nBusyBox v1.19.4 (2012-07-19 22:27:43 CEST) built-in shell (ash)\r\n
Enter 'help' for a list of built-in commands.\r\n\r\n~ # ""
I wanted to know if there is any method in Python as we have in PERL as $telnet->cmd ("cd //tmp/media/..)
Any suggestions are welcomed if you can give an example!!!
You should try to login to the machine using telnet, then you will notice you will login into BusyBox. That string you print not an error it is hte normal BusyBox prompt.
It might not be what you expected, I only know BusyBox from Linux boxes that were unable to properly boot.
I've just setup my EC2 server following this video here exactly - http://www.youtube.com/watch?v=bBajLxeKqoY
I even chose the same server type, everything went well until it asked for the root password to my EC2 server...
Any suggestions?
Update
Updating this answer because of the activity:
Depending on if the system is ubuntu or Rhel the user varies.
For ubuntu it is
ssh -i my-pem-file.pem ubuntu#my-ec2-instance-address
For RHEL it is
ssh -i my-pem-file.pem root#my-ec2-instance-address
Connecting to an ec2 instance does not require a password, it would require only a pem file
and this is how you connect to it
ssh -i my-pem-file.pem ec2-user#my-instance-address
and remember to chmod 400 your pem file before ssh'ing
If you need to do things as root once you are in as ec2-user, use sudo su - that gets you to root and doesn't need a password. Somethings you do need that for, like looking at the tomcat log files
I had the same problem and after a lot of struggle, I read this page again:
http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/AccessingInstancesLinux.html
Notice in the 4th topic it talks about the user name. It is usually ec2-user but if you are using RHEL5 distro it can be root and if you are using the Ubuntu the user name will be ubuntu - which was my case.
So alternatively try those:
ssh -i my-pem-file.pem ubuntu#my-ec2-instance-address for Ubuntu
or
ssh -i my-pem-file.pem root#my-ec2-instance-address for RHEL5
Hope it helps!
You will be asked for password when you enable PasswordAuthentication yes in your sshd_config. Try changing that to no. that should fix it.
Just for anyone else that might have the same problem
Just do
sudo su -
Worked for me
If you've moved/copied the pem file from another machine, the owner of the file may be different to the user that is running the ssh client. To change the owner of the file:
sudo chown <currentusername> <filename.pem>
Please use --query 'KeyMaterial' while generating key.
Info:
keyMaterial - an unencrypted PEM encoded RSA private key.
In my case, we had been copying a text file to keep a ubiquitous key. Someone accidentally added characters to this file, and we started getting prompted for a password from the now-corrupt .pem file, when we had never set a password.
In my case, the problem was the file's break type. Try this:
1.- Open the .pem file with TextWrangler
2.- At Bottom of app, verify if the Break Type is "Windows(CRLF)".
Regards
This is the way to connect:
ssh -i /path/my-key-pair.pem ec2-user#public-ip
Now, instead of ec2-user, it could be root, or centos, or ubuntu, or something else. You can check under the "Usage Instructions". If it's not there, and you've tried all the above users, find out from the documentation that came with the AMI.
There is one bug on AWS that cost me a lot of time. If you're launching an instance from a saved AMI under "Images > AMIs", note that it uses the original keys of the running instance it was created from. When launching the AMI, it will prompt you to choose a new key, and even show such key under the description, but the truth is those keys will never work!
So if you're using a custom/saved AMI you'll have to either get the original keys and use them, or just create a brand new one from AWS or the market place. If you haven't created the AMI or can't remember where it came from, look under the details tab. You can then launch a new instance or create a new AMI from the same source. This will then use the keys that you specify.
chmod 400 pem_file
ssh -i /path_to_the_pem_file ec2-user#ServerIP
or
ssh -i /path_to_the_pem_file ubuntu#ServerIP
root access is prohibited by default by AWS AMIs.
hope it helps.
Check your .pem file is not corrupt. Using this line
openssl rsa -check -in test.pem -noout
It should return RSA Key OK. If it doesn't then there is most likely a problem with some sort of formatting or something in your pem file.
You should use the *.pem instead of a root password. Once you have logged in use passwd to set a password.
It could be a Linux problem.
But there is also a chance that you use the wrong address/key (you started a new instance but still using the old address; or 2 pem have similar name, used the wrong one), "ssh with non-existing user".
I am using TurboLinux instance, on Mac OS system, please try this command:
ssh -i xxxxx.pem root#xxx.xxx.xxx.xxx (public address of allocated VM)
I can get through without asking password anymore.
In case someone else bumps into this, the solution for my problem was that I had to run it with sudo:
sudo ssh -i my-pem-file.pem root#my-ec2-instance-address
For me, the issue was that I had created an AMI from an existing instance, so when I launched the AMI, even though I selected a new key, the instance's original key was what worked.
Login to Amazone console in browser:
https://us-east-2.console.aws.amazon.com/ec2/xxxxxxxx
Click on instances from left panel -> then select your instance -> click on connect button at top right
You will see a window open and there will be a button ssh Client
Here you can see exact command to connect with your instance.