Here is a example script that creates a simple "testjail" with iocage. This script may give you some practical hints, especially if you are writing installation scripts for iocage jails.
For example, you can execute commands in the jail with
There is also another way to execute scripted commands in the jail. Take a look at the section that starts with cat << EOF | ...
This technique let's you do fairly advanced iocage jail installations easily directly from your script.
The way I find the JAILROOT in the script, that is, the path to the root directory of the jail in the FreeNAS directory tree, is not the most elegant way. I would like to know some simpler way to figure out the JAILROOT directory.
It is quite easy to split your script file into multiple sections, which you can output anyway you want. This script creates a rc.d script in the jail from some lines at the end of the script file.
The testjail is now created. You can access the command line of the testjail with
and you can delete the testjail with
As in a script file "$0" refers to the script file itself, the following one-liners give you any section you want from your script file. Just make sure that your divider lines are unique, and no other lines end in the dividers. Not even comment lines! Also, your script MUST exit before the divider lines begin.
I hope these examples can help you to write even better installation scripts for the iocage jails.
For example, you can execute commands in the jail with
iocage exec <jail> <command>There is also another way to execute scripted commands in the jail. Take a look at the section that starts with cat << EOF | ...
This technique let's you do fairly advanced iocage jail installations easily directly from your script.
The way I find the JAILROOT in the script, that is, the path to the root directory of the jail in the FreeNAS directory tree, is not the most elegant way. I would like to know some simpler way to figure out the JAILROOT directory.
It is quite easy to split your script file into multiple sections, which you can output anyway you want. This script creates a rc.d script in the jail from some lines at the end of the script file.
Code:
#!/bin/bash
if [ "$(id -u)" != "$(id -u root)" ]; then
echo "You must be root to run this script."
exit 1
fi
# Let's create a testjail
iocage create -n testjail -r "11.1-RELEASE"
# Let's execute commands in the testjail.
# Everything between the cat-line and the line starting
# with EOF will be executed inside the testjail
cat << EOF | iocage console testjail
echo "Hello from"
hostname
EOF
# The root directory of the testjail can be found at
JAILROOT="$(iocage get jail_zfs_dataset testjail 2>/dev/null | sed -rn 's/(.*)\/data$/\1\/root/p')"
[ -n "$JAILROOT" ] && JAILROOT="/mnt/$JAILROOT"
# Now JAILROOT is the root directory of the jail
# or it is "" if the root directory could not be determined.
if [ -d "$JAILROOT" ]; then
# Let's create a dummy rc.d script in the testjail.
# The contents of the rc.d file is located near
# the end of THIS file after the divider line.
sed '1,/--divider--$/d' "$0" > "$JAILROOT/etc/rc.d/dummy"
chmod 555 "$JAILROOT/etc/rc.d/dummy"
# or alternatively:
# iocage exec testjail "chmod 555 /etc/rc.d/dummy"
# Let's start the dummy service in the testjail
iocage exec testjail "service dummy start"
fi
# This script must end before the divider line.
exit 0
# After the divider line are the contents of a dummy rc.d script
--divider--
#!/bin/sh
. /etc/rc.subr
name=dummy
start_cmd=dummy_start
stop_cmd=":"
dummy_start() {
echo "Hello, I'm a dummy service"
}
load_rc_config $name
run_rc_command "$1"The testjail is now created. You can access the command line of the testjail with
iocage console testjailand you can delete the testjail with
iocage destroy testjailAs in a script file "$0" refers to the script file itself, the following one-liners give you any section you want from your script file. Just make sure that your divider lines are unique, and no other lines end in the dividers. Not even comment lines! Also, your script MUST exit before the divider lines begin.
Code:
#!/bin/sh
# Lines before div1 i.e. the script itself
sed '/--div1--$/,$d' "$0"
# Lines between div1 and div2
awk '/--div1--$/{flag=1;next}/--div2--$/{flag=0}flag' "$0"
# Lines after div2
sed '1,/--div2--$/d' "$0"
# Must exit before the divs
exit 0
--div1--
These lines are
between div1 and div2
--div2--
This line comes after div2,
and so does this.
I hope these examples can help you to write even better installation scripts for the iocage jails.
