Difference between revisions of "Installation on Slackware 14"

Revision as of 13:26, 9 July 2016

The purpose of this note is to setup a guest based on Slackware. What follows was tested on Slackware 14.2 to 13.1 (both 32b and 64b). I will assume that you have a Linux-Vserver box working.

Download

wget http://notes.sagredo.eu/sites/notes.sagredo.eu/files/linux-vserver/slack_vserver.tar.gz
tar xzf slack_vserver.tar.gz
cd slack_vserver
ls

You have downloaded the following files:

• PKG_LIST is the package list to be installed in the guest
• download_slack_pkg.sh is a script that you can use to download all the PKG_LIST. If you use this scriptlet all the package will be stored in the slackware[64]-version_pkg sub-folder.
• make_slack_vserver.sh is the shell script that you have to adjust. It installs the guest.
• rc is the startup script for the virtual server. It will be automatically copied in /etc/rc.d* /init.d/rc
• linux-vserver_slackware-14.2.patch is the patch which modify rc.0, rc.6 rc.M, rc.S, rc.inet2 and rc.syslog. It must be applyed after the creation of the vserver process. In the patch I switched off all the mounts and executable that are related to the hardware.

Download the packages

First of all select a minimal set of packages to be installed on the virtual server. This list of 123 packages is based on the Minimal System reported at http://slackwiki.org/Minimal_System without all hardware, kernel and multimedia related packages. The install leads to a guest of about 460M of size. This set fits with the installation of a complete virtual web server including apache, apache-tomcat, php, mysql, postgresql, qmail and related, ftp, named.

I assume that the packages to be installed are stored in the slackware{$ARCH}-{$VERSION}_pkg folder. If not, adjust its location editing the make_slack_vserver.sh script.

You can download my minimal set of packages running the shell script download_slack_pkg.sh. It can create a folder like slackware{$ARCH}-{$VERSION}_pkg for you, where $ARCH has to be "64" if you want to download 64b packages or empty otherwise, while $VERSION is the Slackware version, so it's something like "14.2".

#!/bin/bash
#
# v. 2016.06.08
# Now the script parses comments in the package list (thanks to Mark Colclough)

VERSION="14.2" # Slackware version
ARCH="64" # you can put 64 for 64b cpu just to separate 64/32 download folders

# Put here your favourite Slackware repository
SRC="ftp://ftp.slackware.no/slackware/slackware${ARCH}-${VERSION}/"

# put here your pkg list
LIST="${PWD}/PKG_LIST"

# the directory where you unpacked slack_vserver.tar.gz
# $PWD should work, otherwise put /path/to/slack_vserver
SETUP=$PWD

# the directory where you want to download the slackware packages
PACKAGES="${SETUP}/slackware${ARCH}-${VERSION}_pkg"

# create the folder where the pkg will be downloaded
if [ ! -d "$PACKAGES" ]; then
        mkdir -p $PACKAGES
fi

# create the "patches" sub-folder
if [ ! -d "${PACKAGES}/patches" ]; then
        mkdir -p "${PACKAGES}/patches"
fi

# download
cd $PACKAGES

if [ -f $LIST ]; then
        while read LINE
            do
            [ "$LINE" ] || continue
            [ "${LINE#\#}" = "$LINE" ] || continue
            wget "${SRC}slackware${ARCH}/${LINE}*.t?z"
        done < $LIST
else
        echo "Can't find $LIST file."
        exit 1
fi

# download packages from the patches folder
cd ${PACKAGES}/patches

if [ -f ${LIST} ]; then
        while read LINE
        do
                IFS='/' read -ra PKG <<< "$LINE"
                [ "${PKG#\#}" = "${PKG}" ] || continue
                PKG_LEN=${#PKG[@]}
                if [ $PKG_LEN == 2 ]; then
                        wget "${SRC}patches/packages/${PKG[1]}*.t?z"
                fi
        done < $LIST
else
        echo "Can't find $LIST file."
        exit 1
fi

NB: this script tries also to overwrite the packages downloaded from the /slackware folder with the updates belonging to the /patches dir.

Make the guest

Now let's create the guest and install the packages. As you know you must choose at least a "name", a "context" and an ip. In addition you have to modify most of the rc.* startup scripts removing all the hardware related daemons, and finally replace the /dev dir.

This is done adjusting and running the script make_slack_vserver.sh:

#!/bin/bash
#
# v. 2016.07.05
# Author: Roberto Puzzanghera
# Thanks to Mark Colclough for corrections
#
# This script installs a Slackware guest into a linux-vserver host (http://linux-vserver.org)
#
# Comments are welcome :-)
# More info here: http://notes.sagredo.eu/node/7

# adjust this to where your things live
NAME=test
HOSTNAME=$NAME.YOURDOMAIN.XY
IP=10.0.0.182
INTERFACE=eth0:$IP/24
CONTEXT=5182
VERSION=14.2 # Slackware version
ARCH="64" # you can put 64 for 64b cpu just to separate 64/32 download folders

# where is the vservers dir? default is /vservers
VDIR="/usr/local/vservers"

# the directory where you unpacked slack_vserver.tar.gz
# $PWD should work, otherwise put /path/to/slack_vserver
SETUP=$PWD

# the directory where you downloaded the slackware packages
PACKAGES="${SETUP}/slackware${ARCH}-${VERSION}_pkg"

# the path to rc script file (leave as is)
RC="${SETUP}/rc"

################### end configuration

# sanity check

if [ ! -d "$VDIR" ]; then
        echo
        echo "Can't find VDIR dir: $VDIR"
        echo "Exiting"
        echo
        exit 1
fi
if [ ! -d "$SETUP" ]; then
        echo 
        echo "Can't find SETUP dir: $SETUP"
        echo "Exiting"
        echo
        exit 1
fi
if [ ! -d "$PACKAGES" ]; then
        echo
        echo "Can't find PACKAGES dir: $PACKAGES"
        echo "Exiting"
        echo
        exit 1
fi
if [ ! -f "$RC" ]; then
        echo
        echo "Can't find RC path: $RC"
        echo "Exiting"
        echo
        exit 1
fi

# if everything is ok start the install

echo
read -p "press any key to make skeleton..."
vserver ${NAME} build -m skeleton \
        --hostname ${HOSTNAME} \
        --interface ${INTERFACE} \
        --context $CONTEXT \
        --flags lock,virt_mem,virt_uptime,virt_cpu,virt_load,sched_hard,hide_netif \
        --initstyle sysv
echo "...done"
echo

read -p "press any key to move the /dev folder to a temp dir the /dev folder..."
mv $VDIR/$NAME/dev $VDIR/$NAME/dev2

echo
read -p "press any key to install packages..."
cd $PACKAGES
installpkg --root $VDIR/$NAME *.t?z;
ROOT=$VDIR/$NAME upgradepkg patches/*.t?z;
echo "...done"
echo

echo
read -p "press any key to copy the rc script to /etc/rc.d/init.d..."
echo
echo "copying rc to /etc/rc.d/init.d/rc"
cp -p $RC $VDIR/$NAME/etc/rc.d/init.d/
echo "...done"

echo
echo "removing x flag to rc.sshd and rc.inetd, removing not needed rc scripts"
chmod -x $VDIR/$NAME/etc/rc.d/rc.sshd $VDIR/$NAME/etc/rc.d/rc.inetd
rm $VDIR/$NAME/etc/rc.d/rc.cpufreq $VDIR/$NAME/etc/rc.d/rc.modules* $VDIR/$NAME/etc/rc.d/rc.inet1* $VDIR/$NAME/etc/rc.d/rc.loop
echo "...done"
echo

echo "trying to adjust HOSTNAME, hosts, resolv.conf, profile. Check them later..."
cp /etc/resolv.conf $VDIR/$NAME/etc/
cp /etc/localtime $VDIR/$NAME/etc/
rm $VDIR/$NAME/etc/profile
cp /etc/profile $VDIR/$NAME/etc/
echo $HOSTNAME >  $VDIR/$NAME/etc/HOSTNAME
echo "127.0.0.1 localhost" > $VDIR/$NAME/etc/hosts
echo "$IP $HOSTNAME $NAME" >> $VDIR/$NAME/etc/hosts
touch $VDIR/$NAME/etc/mtab
touch $VDIR/$NAME/etc/fstab
echo "...done"
echo

read -p "press any key to restore /dev2 to /dev"
rm -r $VDIR/$NAME/dev
mv $VDIR/$NAME/dev2 $VDIR/$NAME/dev

echo
echo -n "Do you want that I apply the patch for you y/n? [y] "
read VAR_PATCH

if [ "$VAR_PATCH" = 'y' ] || [ "$VAR_PATCH" = '' ]; then

        if [ ! -f "${SETUP}/linux-vserver_slackware-${VERSION}.patch" ]; then
                echo
                echo "Can't find any PATCH here: ${SETUP}/linux-vserver_slackware-${VERSION}.patch"
                echo "Exiting"
                echo
                exit 1
        fi

        cd ${VDIR}/${NAME}/etc/rc.d
        patch -p1 < ${SETUP}/linux-vserver_slackware-${VERSION}.patch
        echo "patch applyed."
        echo
        echo "You can start and enter the virtual server typing: "
        echo
        echo "vserver $NAME start"
        echo "vserver $NAME enter"

        else
                echo
                echo "DON'T FORGET to patch /etc/rc.d as follows: "
                echo
                echo "cd $VDIR/$NAME/etc/rc.d"
                echo "patch -p1 < $SETUP/linux-vserver_slackware-$VERSION.patch"
fi

echo
echo "More info on http://notes.sagredo.eu/node/7"
echo

Apply the patch

The script itsself can install the patch for you. Anyway this is how to apply the patch by yourself:

cd /vservers/vserver_name/etc/rc.d
patch -p1 < /path/to/slack_vserver/slackware-14.2.patch

Inside the tarball you can find the old patch for Slackware 14.1, 14.0, 13.37 and 13.1, if you like.

Post installation tasks

Put this inside your rc.local:

/usr/local/etc/rc.d/init.d/vprocunhide  start

Start the new server

vserver <vserver_name> start
vserver <vserver_name> enter

Patch applayed

diff -ruN rc.d-original/rc.6 rc.d/rc.6
--- rc.d-original/rc.6	2016-03-26 17:48:37.000000000 +0100
+++ rc.d/rc.6	2016-07-09 14:08:48.470713168 +0200
@@ -37,36 +37,6 @@
 		;;
 esac
 
-# Save the system time to the hardware clock using hwclock --systohc.
-# This will also create or update the timestamps in /etc/adjtime.
-if [ -x /sbin/hwclock ]; then
-  # Check for a broken motherboard RTC clock (where ioports for rtc are
-  # unknown) to prevent hwclock causing a hang:
-  if ! grep -q " : rtc" /proc/ioports ; then
-    CLOCK_OPT="--directisa"
-  fi
-  if [ /etc/adjtime -nt /etc/hardwareclock ]; then
-    if grep -q "^LOCAL" /etc/adjtime ; then
-      echo "Saving system time to the hardware clock (localtime)."
-    else
-      echo "Saving system time to the hardware clock (UTC)."
-    fi
-    /sbin/hwclock $CLOCK_OPT --systohc
-  elif grep -q "^UTC" /etc/hardwareclock 2> /dev/null ; then
-    echo "Saving system time to the hardware clock (UTC)."
-    if [ ! -r /etc/adjtime ]; then
-      echo "Creating system time correction file /etc/adjtime."
-    fi
-    /sbin/hwclock $CLOCK_OPT --utc --systohc
-  else
-    echo "Saving system time to the hardware clock (localtime)."
-    if [ ! -r /etc/adjtime ]; then
-      echo "Creating system time correction file /etc/adjtime."
-    fi
-    /sbin/hwclock $CLOCK_OPT --localtime --systohc
-  fi
-fi
-
 # Run any local shutdown scripts:
 if [ -x /etc/rc.d/rc.local_shutdown ]; then
   /etc/rc.d/rc.local_shutdown stop
@@ -126,10 +96,6 @@
   sleep $FUSER_DELAY
 fi
 
-# Unmount any NFS, SMB, or CIFS filesystems:
-echo "Unmounting remote filesystems:"
-/bin/umount -v -a -l -f -r -t nfs,smbfs,cifs | tr -d ' ' | grep successfully | sed "s/:successfullyunmounted/ has been successfully unmounted./g"
-
 # Try to shut down pppd:
 PS="$(ps ax)"
 if echo "$PS" | /bin/grep -q -w pppd ; then
@@ -215,22 +181,11 @@
 # Before unmounting file systems write a reboot or halt record to wtmp.
 $shutdown_command -w
 
-# Turn off swap:
-echo "Turning off swap."
-/sbin/swapoff -a
-/bin/sync
-
 # Stop cgmanager and cgproxy:
 if [ -x /etc/rc.d/rc.cgmanager ]; then
   sh /etc/rc.d/rc.cgmanager stop
 fi
 
-echo "Unmounting local file systems:"
-/bin/umount -v -a -t no,proc,sysfs | tr -d ' ' | grep successfully | sed "s/:successfullyunmounted/ has been successfully unmounted./g" 2> /dev/null
-
-echo "Remounting root filesystem read-only:"
-/bin/mount -v -n -o remount,ro /
-
 # This never hurts:
 /bin/sync
 
@@ -288,12 +243,3 @@
     fi
   fi
 fi
-
-# Now halt (poweroff with APM or ACPI enabled kernels) or reboot.
-if [ "$shutdown_command" = "reboot" ]; then
-  echo "Rebooting."
-  /sbin/reboot
-else
-  /sbin/poweroff
-fi
-
diff -ruN rc.d-original/rc.M rc.d/rc.M
--- rc.d-original/rc.M	2016-05-05 06:27:00.000000000 +0200
+++ rc.d/rc.M	2016-07-09 14:14:20.090096570 +0200
@@ -20,10 +20,6 @@
   /sbin/ldconfig &
 fi
 
-# Screen blanks after 15 minutes idle time, and powers down in one hour
-# if the kernel supports APM or ACPI power management:
-/bin/setterm -blank 15 -powersave powerdown -powerdown 60
-
 # Set the hostname.
 if [ -r /etc/HOSTNAME ]; then
   /bin/hostname $(cat /etc/HOSTNAME | cut -f1 -d .)
@@ -109,13 +105,6 @@
   sh /etc/rc.d/rc.bluetooth start
 fi
 
-# Start wicd or networkmanager:
-if [ -x /etc/rc.d/rc.wicd -a -x /usr/sbin/wicd ]; then
-  sh /etc/rc.d/rc.wicd start
-elif [ -x /etc/rc.d/rc.networkmanager ]; then
-  sh /etc/rc.d/rc.networkmanager start
-fi
-
 # Start networking daemons:
 if [ -x /etc/rc.d/rc.inet2 ]; then
   . /etc/rc.d/rc.inet2
@@ -126,14 +115,6 @@
   . /etc/rc.d/rc.scanluns
 fi
 
-# Mount any additional filesystem types that haven't already been mounted:
-mount -a -v 2> /dev/null | grep -v -e "already mounted" -e "ignored" | cut -f 1 -d : | tr -d ' ' | while read dev ; do mount | grep "${dev} " ; done
-
-# Start the Control Script for automounter:
-if [ -x /etc/rc.d/rc.autofs ]; then
-  sh /etc/rc.d/rc.autofs start
-fi
-
 # Start the Network Time Protocol daemon:
 if [ -x /etc/rc.d/rc.ntpd ]; then
   sh /etc/rc.d/rc.ntpd start
@@ -153,16 +134,6 @@
 chmod 755 / 2> /dev/null
 chmod 1777 /tmp /var/tmp
 
-# Start ACPI daemon.
-if [ -x /etc/rc.d/rc.acpid ]; then
-  . /etc/rc.d/rc.acpid start
-fi
-
-# Enable CPU frequency scaling:
-if [ -x /etc/rc.d/rc.cpufreq ]; then
-  . /etc/rc.d/rc.cpufreq start
-fi
-
 # Update any existing icon cache files:
 if find /usr/share/icons -maxdepth 2 2> /dev/null | grep -q icon-theme.cache ; then
   for theme_dir in /usr/share/icons/* ; do
diff -ruN rc.d-original/rc.inet2 rc.d/rc.inet2
--- rc.d-original/rc.inet2	2012-08-15 00:53:35.000000000 +0200
+++ rc.d/rc.inet2	2016-07-09 14:10:49.388841574 +0200
@@ -16,51 +16,6 @@
 # At this point, we are ready to talk to The World...
 
 
-# Mount remote (NFS) filesystems:
-if cat /etc/fstab | grep -v '^#' | grep -w nfs 1> /dev/null 2> /dev/null ; then
-  # Start rpc.portmap, /sbin/rpc.lockd, and /sbin/rpc.statd if we find NFS
-  # volumes defined in /etc/fstab since these will need to be running in order
-  # to mount them.  If they are not running, attempting to mount an NFS
-  # partition will cause mount to hang, or at least result in unreliable
-  # operation.  Keep this in mind if you plan to mount unlisted NFS
-  # partitions... 
-  # If you have uncommented NFS partitions in your /etc/fstab, rc.rpc is run
-  # whether it is set as executable or not.  If you don't want to run it,
-  # comment the NFS partitions out in /etc/fstab or erase/rename rc.rpc.
-  if [ -r /etc/rc.d/rc.rpc ]; then
-    sh /etc/rc.d/rc.rpc start
-  fi
-  echo "Mounting remote (NFS) file systems:  /sbin/mount -a -t nfs"
-  /sbin/mount -a -t nfs          # This may be our /usr runtime!
-  # Show the mounted volumes:
-  /sbin/mount -v -t nfs
-fi
-
-# If /etc/rc.d/rc.rpc is executable, run it to load rpc.portmap, rpc.lockd,
-# and rpc.statd.  This might be needed to mount NFS partitions that are not
-# listed in /etc/fstab.  Starting this twice won't hurt as the script will
-# check if things are already running before trying to start them.
-if [ -x /etc/rc.d/rc.rpc ]; then
-  sh /etc/rc.d/rc.rpc start
-fi
-
-# Mount remote CIFS filesystems.  Note that where possible, using CIFS is
-# preferred over SMBFS.  SMBFS is no longer actively maintained.
-if cat /etc/fstab | grep -v '^#' | grep -w cifs 1> /dev/null 2> /dev/null ; then
-  echo "Mounting remote CIFS file systems:  /sbin/mount -a -t cifs"
-  /sbin/mount -a -t cifs
-  # Show the mounted volumes:
-  /sbin/mount -v -t cifs
-fi
-
-# Mount remote SMB filesystems:
-if cat /etc/fstab | grep -v '^#' | grep -w smbfs 1> /dev/null 2> /dev/null ; then
-  echo "Mounting remote SMBFS file systems:  /sbin/mount -a -t smbfs"
-  /sbin/mount -a -t smbfs
-  # Show the mounted volumes:
-  /sbin/mount -v -t smbfs
-fi

Contact

Comments and criticism can be addressed to roberto dot puzzanghera at sagredo dot eu (http://notes.sagredo.eu/node/7)