ec2 run-instances user-data script - changing user from root to ubuntu - linux

I am using aws cli to launch ec2 instances. I am using the user-data parameter to run a custom script when they are launched:
aws ec2 run-instances \
......
--user-data file://~/Desktop/script.sh
In script.sh i can see that it is running it user root user
for example when i run
whoami > testwhoami.txt
i can see root in the text file
I need to switch the user to ubntu and it doesn't seem to work.
I have tried many things in this script:
sudo su -l ubuntu
su -l ubuntu &> output.txt
but after all these whoami keeps showing that root is the user and output.txt is empty as well
What might cause this? How can i debug it properly?

Your command su -l ubuntu returns an interactive shell, which is useless when you're running a script (EC2 user-data).
You need to submit your command using su -c option like this:
su ubuntu -c 'whoami' > output.txt

Related

Linux shell script - How to switch user and run a script?

I'm currently writing a .sh script to deploy different applications on 4 different machines. Right now I'm having trouble with running a script as another user. I need to log in with myUser with my credentials and then sudo su to user2 to run a specific script.
Normally, I would manually do the following:
ssh myUser#remotehost
[Type in password]
sudo su - user2
cd /path/only/accessible/to/user2
./someScript.sh
when I tried
ssh -t myUser#$remotehost "sudo su - user2 && /path/only/accessible/to/user2 && ./someScript.sh"
I was asked my password, then stayed logged as user2, without any feedback from the script, which would normally give me some informations.
What am I doing wrong?
Try
ssh -t myUser#$remotehost "sudo -u user2 /path/only/accessible/to/user2/someScript.sh"
If you need shell access after that you can use
ssh -t myUser#$remotehost "sudo -u user2 /path/only/accessible/to/user2/someScript.sh && /bin/bash -l"
An update if anyone wonders about this.
What I finally did was to log in with an ssh key. My sysadmin had to get involved in order to set it up, but at least it is a viable option.
ssh -i /path/to/sshKey user2#$remoteHost "/path/only/accessible/to/user2/someScript.sh"

Not able to login as jenkins user

I have jenkins installed on amazon ec2 instance but i am not able to login as jenkins user.I used this command
sudo su - jenkins
but it is taking me to bash instead of logging me as a jenkis user
You are logged in as jenkins user, as you verified using whoami.
The bash prompt for this user is just not configured to include the user name, but only displays the shell version.
Try the following:
[root#ip-172-20-0-211 ~]# echo $PS1
\[\u#\h \w .... some more cryptic stuff
Copy this output and paste it here
[root#ip-172-20-0-211 ~]# sudo su - jenkins
-bash-4.2$ export PS1="\[\u#\h \w .... some more cryptic stuff"
and the next line should look similar to the following:
[jenkins#ip-172-20-0-211 ~]
Read more about the prompt e.g. here.
If you want to log in as Jenkins, you can do so with
sudo su -s /bin/bash jenkins

bash & s3cmd not working properly

Hi I have a shell script which contains s3cmd command on ubuntu 12.04 LTS.
I configured cron for this shell script which works fine for local environment but don't push the file to s3. But when i run shell script manually, It pushes the file to s3 without any error. I checked log and found nothing for this. Here is my shell script.
#!/bin/bash
User="abc"
datab="abc_xyz"
pass="abc#123"
Host="abc1db.instance.com"
FILE="abc_rds`date +%d_%b_%Y`.tar.gz"
S3_BKP_PATH="s3://abc/db/"
cd /abc/xyz/scripts/
mysqldump -u $User $datab -h $Host -p$pass | gzip -c > $FILE | tee -a /abc/xyz/logs/app-bkp.log
s3cmd --recursive put /abc/xyz/scripts/$FILE $S3_BKP_PATH | tee -a /abc/xyz/logs/app-bkp.log
mv /abc/xyz/scripts/$FILE /abc/xyz/backup2015/Database/
#END
This is really weird. Any suggestion would be a great help.
Check if the user running configured in crontab has correct permissions and keys in the environment.
I am guessing the keys are configured in env file as they are not here in the script.

rundeck - switch to root user in job script

Logging via terminal I can switch to root user fine:
ubuntu#ip-10-0-0-70:~$ sudo -s
root#ip-10-0-0-70:~# whoami
root
So I created in rundeck a job script with this:
whoami;
echo "1st step";
sudo -s;
echo "2nd step";
And when I run this, it prints:
ubuntu
1st step
After print '1st step' it get stucked forever. Seems a problem with sudo -s command.
tried sudo -i but the same happens
tried sudo su - root but the same happens
rundeck is logging as ubuntu user, me too
any idea to switch to root in rundeck script?
This is the expected behaviour.
You are running a shell via 'sudo -s' and then not leaving/exiting it ! So it waits forever for somethig that won't come.
You can probably add 'sudo' as an Advanced option of your script (where it says "Run script with an interpreter or prefix. E.g.: sudo, time:").
But it will run your whole script as root.
If you just want a specific command to be run as root , just prefix your command with sudo as so:
sudo "enter_your_command_to_be_run_as_root_here"
Entering the command prefixed by Sudo will generate the following error on most linux distributions.
sudo: sorry, you must have a tty to run sudo
You can enable sudo without tty by running 'visudo' and commenting out the defaults line or removing 'requiretty' from the defaults line.
Details can be found here:
http://www.cyberciti.biz/faq/linux-unix-bsd-sudo-sorry-you-must-haveattytorun/

ssh command to run remote script exist shell on remote server when switching user

When I run a script such as this:
ssh -t root#10.10.10.10 '/tmp/somescript.sh'
where the script is defined as:
#!/bin/sh
mkdir -p /data/workday/cred
chown -R myuser:myuser /data
su myuser - # <------- NOTICE THIS ! ! ! !
rpm -Uvp --force --nodeps --prefix /data/place /data/RPMs/myrpm.rpm
Notice the above su command.
If I comment-out the su command, the script runs remotely and then my shell prompt returns to where I came from ( same server where I ran the ssh command above )
But leaving the script as listed above, causes the script to complete successfully but the shell prompt stays on the remote server.
How can I prevent that ? Making sure that the issuer of the rpm command is a different user than root just a listed ?
But leaving the script as listed above, causes the script to complete successfully but the shell prompt stays on the remote server.
Not exactly. The script is running up to the su command, which spawns a new subshell, and stopping there until you exit the shell. Until you exit that shell, the rpm command never runs, and when it does, it runs as root.
If you want to run the rpm command as a non-root user, you'd need to do something a little different, like:
sudo -u myuser rpm -Uvp ...
add 'exit' at the end of your script

Resources