HowTo: Change disk checking/fsck at boot frequency

[Old Pink]

Hi,

Nearly all Ubuntu users know, that every 30 boots their system performs what is know as an "fsck" to make sure your hard drive has no errors.

If you go through periods of frequent rebooting, these checks can become annoying.

To change how often these boot up checks occur, run this command in a terminal:

sudo tune2fs -c 50 /dev/hda1

The two things in bold can be modified.

First, the number "50", which chooses how often you want the check performed. 1 makes it scan at every boot, 0 stops scanning altogether, and all other numbers mean the scan will be performed every ** boots, 20 for 20, 30 for 30, 40 for 40 and so on.

Next, the "hda1" which you simply change to match the location of your hard drive, be it hda1,2,3 or even sda1,2,3.

Setting the scan to never happen isn't recommended, I'd say between 50-100 is safe.

- Matt

[bodhi.zazen]

Nice How-to Old Pink

This thread has been added to the UDSF wiki.

Change Disk Check Frequency

I took the liberty of adding some basic information about fstab:

fstab options
/etc/fstab is a system configuration file and is used to tell the Linux kernel which partitions (file systems) to mount and where on the file system tree.

A typical fatab entry may look like this:

/dev/hda1 / ext3 defaults 1 1
/dev/hdb1 /home ext3 defaults 1 2

The 6th column (in bold) is a fsck options.

• 0 = Do not check.
• 1 = First file system (partition) to check;
  
  • / (root partition) should be set to 1.
• 2 = ALL OTHER file systems to be checked.

bodhi.zazen

[Charles Hand]

Short of turning it off completely (not recommended) it's still going to happen by surprise and on the computer's timetable. Is there any way to tie into the mechanism that counts the boots and run a little program that says "you're going to have a forced check three boots from now"? Maybe a log file that tells how may boots are left, which could be grepped?

[Sam]

There is also the -i switch to set an interval before the check.

-i interval-between-checks[d|m|w]

Example combining both methods:

$ sudo tune2fs -c 50 -i 1m /dev/hda1

[Charles Hand]

Short of turning it off completely (not recommended) it's still going to happen by surprise and on the computer's timetable. Is there any way to tie into the mechanism that counts the boots and run a little program that says "you're going to have a forced check three boots from now"? Maybe a log file that tells how may boots are left, which could be grepped?

Here it is:

sudo dumpe2fs -h /dev/sdb1 | grep -i 'mount count'

Output:

dumpe2fs 1.38 (30-Jun-2005)
Mount count:              5
Maximum mount count:      30

[kvonb]

Search the forums for "bonager", it is a panel app that warns you of impending scans and lets you postpone them

[MaineDave]

There's also a showfsck package (I'm using Edgy in case that matters) with the command "showfsck" to show how many reboots remain before a fsck runs.

[lunamystry]

how do i get the Ubuntuuser icon for my website?

[NuK]

I'm learning bash and I wrote this script that shows the "mount count" for each partition. It works fine for me... I hope it helps!

#!/bin/bash

# this directory must exist to execute the script
if [ -d /dev/disk/by-id/ ]; then :; else echo 'ERROR: /dev/disk/by-id/ does not exist'; exit 1; fi

# root user is necessary
if [ $UID -gt 0 ]; then	echo 'ERROR: You have to be root'; exit 1; fi

# list ALL partitions and exclude the SWAP ones
declare -a SWAPS=( $(cat /proc/swaps | grep /dev/ | awk '{print$1}') )
lsParts="ls -l /dev/disk/by-id/ | grep --regexp=-part[0-9] | awk '{sub(\"../../\",\"/dev/\");"
for i in $(seq 0 $((${#SWAPS[@]} - 1))); do
	lsParts=${lsParts}' sub("'${SWAPS[$i]}'", "");'
done
lsParts=${lsParts}" print\$10}'"
declare -a PARTITIONS=(`eval $lsParts`)

# format TEMP file
for i in $(seq 0 $((${#PARTITIONS[@]} - 1))); do
	PARTITION=${PARTITIONS[$i]}
	if [ -n $PARTITION ]; then
		# first line, device name
		echo $PARTITION >> .temp.$$
		# mount info
		sudo dumpe2fs -h $PARTITION | grep 'Filesystem volume name' >> .temp.$$
		sudo dumpe2fs -h $PARTITION | grep 'Mount count' >> .temp.$$
		sudo dumpe2fs -h $PARTITION | grep 'Maximum mount count' >> .temp.$$
		sudo dumpe2fs -h $PARTITION | grep 'Last mount time' >> .temp.$$
		sudo dumpe2fs -h $PARTITION | grep 'Last checked' >> .temp.$$
		echo '-------------------------------' >> .temp.$$
	fi
done

# show and delete TEMP file
clear
cat .temp.$$
rm -f .temp.$$

# TUNE help
echo
echo 'HOWTO'
echo -e "- Change Filesystem volume name:\n\tsudo tune2fs -L NewLabel /dev/hda1\n\tsudo e2label /dev/hda1 NewLabel"
echo
echo -e "- Change Maximum mount count:\n\tsudo tune2fs -c 50 /dev/hda1"
echo
echo -e "- Modify reserved space:\n\tsudo tune2fs -m 1 /dev/hda1"
echo

exit 0

Remember that you can call the script like this:

sudo bash ./scriptname.sh

But if you want to use it more than once, it's better to change the execute permissions and call it without the "bash" command:

chmod 755 ./scriptname.sh
sudo ./scriptname.sh

[ntlam]

Hi,

Nearly all Ubuntu users know, that every 30 boots their system performs what is know as an "fsck" to make sure your hard drive has no errors.

If you go through periods of frequent rebooting, these checks can become annoying.

To change how often these boot up checks occur, run this command in a terminal:

sudo tune2fs -c 50 /dev/hda1

The two things in bold can be modified.

First, the number "50", which chooses how often you want the check performed. 1 makes it scan at every boot, 0 stops scanning altogether, and all other numbers mean the scan will be performed every ** boots, 20 for 20, 30 for 30, 40 for 40 and so on.

Next, the "hda1" which you simply change to match the location of your hard drive, be it hda1,2,3 or even sda1,2,3.

Setting the scan to never happen isn't recommended, I'd say between 50-100 is safe.

- Matt

Nice guide