PackageKit can’t find file in /var/cache/PackageKit/

After installing RPM Fusion repos on Fedora 28 you might be getting an error shown by update manager every few minutes in GUI.
Error will say something like this: “Error when getting information for file “/var/cache/PackageKit/28/metadata/rpmfusion-nonfree-nvidia-driver/repodata/appstream.xml.gz”: No such file or directory

Error is described here https://bugzilla.redhat.com/show_bug.cgi?id=1580162 and seem to be caused by a repo that claims to support appdata but doesn’t actually provide it.

Error can be resolved by changing enabled_metadata=1 to enabled_metadata=0 in repo description file.
In case of above error change was made in /etc/yum.repos.d/rpmfusion-nonfree.repo and it now looks like this:

[rpmfusion-nonfree]
name=RPM Fusion for Fedora $releasever - Nonfree
#baseurl=http://download1.rpmfusion.org/nonfree/fedora/releases/$releasever/Everything/$basearch/os/
metalink=https://mirrors.rpmfusion.org/metalink?repo=nonfree-fedora-$releasever&arch=$basearch
enabled=1
enabled_metadata=0 #changed from 1 to 0
metadata_expire=14d
type=rpm-md
gpgcheck=1
repo_gpgcheck=0
gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-rpmfusion-nonfree-fedora-$releasever

[rpmfusion-nonfree-debuginfo]
name=RPM Fusion for Fedora $releasever - Nonfree - Debug
#baseurl=http://download1.rpmfusion.org/nonfree/fedora/releases/$releasever/Everything/$basearch/debug/
metalink=https://mirrors.rpmfusion.org/metalink?repo=nonfree-fedora-debug-$releasever&arch=$basearch
enabled=0
metadata_expire=7d
type=rpm-md
gpgcheck=1
repo_gpgcheck=0
gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-rpmfusion-nonfree-fedora-$releasever

[rpmfusion-nonfree-source]
name=RPM Fusion for Fedora $releasever - Nonfree - Source
#baseurl=http://download1.rpmfusion.org/nonfree/fedora/releases/$releasever/Everything/source/SRPMS/
metalink=https://mirrors.rpmfusion.org/metalink?repo=nonfree-fedora-source-$releasever&arch=$basearch
enabled=0
metadata_expire=7d
type=rpm-md
gpgcheck=1
repo_gpgcheck=0
gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-rpmfusion-nonfree-fedora-$releasever

After the change is made run pkcon repair and then pkcon update to check for any errors.

[[email protected] ~]# pkcon repair
                              [=========================]         
Finished                      [=========================]         
                              [=========================]         
Finished                      [=========================]         
[[email protected] ~]# pkcon update
Getting updates               [=========================]         
Finished                      [=========================]         
No packages require updating to newer versions.
[[email protected] ~]# 

Automatic backup of Ubiquiti ES-48-LITE over SSH

Ubiquiti ES-48-LITE currently doesn’t have an option to copy startup config to remote machine over scp.

This script will connect to Ubiquiti EdgeSwitch  over SSH and save startup config locally on machine running the script.

#!/usr/bin/expect -f

set user "username"
set sshpass "password for ssh access"
set enablepass "password for enable mode"
set device "device IP or hostname"
set cfgbackup "/path/to/save/config"

spawn ssh [email protected]$device

expect "assword:"
send "$sshpass\r"

expect ">"
send "enable\r$enablepass\rterminal length 0\r"

expect "#"
log_file -noappend $cfgbackup
send "show startup-config\r"
expect "#\n"
log_file

send "exit\rexit\r"

exec /bin/sed -i {1d;$d} $cfbbackup

 

Find which directory is using most inodes

Oneliners to find which directory on the server is using most inodes:

To find how many inodes are used per each directory on certain path run:

find ./ -maxdepth 1 -type d -print0 | while IFS= read -r -d $'\0' dir; do echo "$(find "$dir" | wc -l) "$dir""; done| sort -n | tail

Last line will show the path being searched and second to last line will be directory with most inodes, then cd to that directory and run again to see which subfolder is using most inodes.

To get directory with most inodes immediately, without the need to cd in each directory you can run following:

find ./ -type d -print0 | while IFS= read -r -d $'\0' dir; do echo "$(stat -t "$dir" | awk '{print$2}') "$dir""; done | sort -n | tail

This will not show the number of inodes used, only which directory has most inodes in it.
Oneliner from top can be used to get exact number of inodes in it.

Small WordPress backup script

Simple backup script written in bash, for WordPress on Linux, that will dump MySQL database used by WordPress site and create tar.gz file consisting of WordPress site files and database dump

#!/bin/bash

BKPDIR=/local/backup/directory
WEBROOT=/wordpress/install/directory/

DBUSER=$(grep DB_USER $WEBROOT/wp-config.php | awk -F\' '{print$4}')
DBNAME=$(grep DB_NAME $WEBROOT/wp-config.php | awk -F\' '{print$4}')
DBPASSWORD=$(grep DB_PASSWORD $WEBROOT/wp-config.php | awk -F\' '{print$4}')
DBDUMP="$BKPDIR""$DBNAME"_$(date +"%Y-%m-%d-%H-%M").sql

#In case you want to rsync backups to remote server
RUSER=remoteuser
RHOST=remoteserver
RDIR=/remote/backup/directory/
RSSHPORT=22

mysqldump -u $DBUSER -p$DBPASSWORD $DBNAME > $DBDUMP

tar -czvf "$BKPDIR"wpbackup_$(date +"%Y-%m-%d_%H-%M").tar.gz $WEBROOT $DBDUMP

rsync -az $BKPDIR -e "ssh -p $RSSHPORT" [email protected]$RHOST:$RDIR

It will create two backup files, one for database dump in format that looks like “joomla_2018-01-04-06-00.sql” and one tar.gz file that will look like “wpbackup_2018-01-04_06-00.tar.gz”.

Script can be saved as wpbackup.sh and put in crontab to run at a certain time like 6:00AM.

0 6 * * * /path/to/script/wpbackup.sh 2>/dev/null

It can also be run manually with bash wpbackup.sh

Ubuntu shows “No Bootable Device” after installation in UEFI mode

When doing a fresh install of Ubuntu 16.04, or any other version of Ubuntu or Linux, after the install you might get your laptop or desktop machine not booting the installed OS, but instead returning “No Bootable Device” message, if laptop is set to UEFI mode.

You might be presented with this image after rebooting your laptop, after a successful OS install.

No Bootable Device message
No Bootable Device message

To resolve this issue reboot your laptop and enter BIOS.

There go to “Select and UEFI file as trusted for executing:” and press Enter key.

BIOS Select UEFI file

Once there choose HDD1.

BIOS select Hard Drive

Then choose <EFI>.

BIOS select EFI

Then you should be able to see <ubuntu> or possibly some other name of your Linux installation.

BIOS select Ubuntu

Then choose shimx64.efi

BIOS select shimx64

And then choose Yes when asked if you want to add this file to allowable database.

BIOS add file to database

Reboot your laptop and you should now be able to boot your OS normally.