Skip to content

Commit

Permalink
shellcheck fixes; --dry-run
Browse files Browse the repository at this point in the history 
Several bunches of fixes:
- Fix all shellcheck (https://github.com/koalaman/shellcheck) warnings
- Add support for --dry-run argument that only echos what would be done
- Add support for listing devices to check in lvcheck.conf
- Add support for passing in list of devices to check from the caller
- Also source an lvcheck.conf in $PWD if available
- Don't check devices whose filesystem TYPE could not be determined
- Support thin-LV devices
- Set max-mount-count on forcing fsck

Signed-off-by: Brian J. Murrell <[email protected]>
  • Loading branch information
@brianjmurrell
brianjmurrell committed Apr 7, 2018
1 parent 890e31c commit f8f5b04
Show file tree
Hide file tree
Showing 2 changed files with 83 additions and 34 deletions.
112 changes: 78 additions & 34 deletions lvcheck
View file
Original file line number Diff line number Diff line change
Expand Up @@ -56,8 +56,8 @@ function log() {
local arg=

# log warning-or-higher messages to stderr as well
[ "$sev" == "emerg" -o "$sev" == "alert" -o "$sev" == "crit" -o \
"$sev" == "err" -o "$sev" == "warning" ] && arg=-s
[ "$sev" == "emerg" ] || [ "$sev" == "alert" ] || [ "$sev" == "crit" ] || \
[ "$sev" == "err" ] || [ "$sev" == "warning" ] && arg=-s

logger -t lvcheck $arg -p user."$sev" -- "$msg"
}
Expand All @@ -75,7 +75,7 @@ function on_ac_power() {
# ignore batteries
[ "${type}" = "Battery" ] && continue

online="`cat "${psu}/online"`"
online="$(cat "${psu}/online")"

[ "${online}" = 1 ] && return 0
[ "${online}" = 0 ] && any_known=yes
Expand Down Expand Up @@ -117,7 +117,7 @@ function try_force_check() {

case "$fstype" in
ext2|ext3|ext4)
tune2fs -C 16000 "$dev"
$ECHO tune2fs -C 16000 -c 15999 "$dev"
;;
xfs)
# XFS does not enforce check intervals; let email suffice.
Expand All @@ -129,13 +129,14 @@ function try_force_check() {
}

# attempt to set the last-check time on $1 to now, and the mount count to 0.
# disable the maximum mount count, since we are now checking periodically.
function try_delay_checks() {
local dev="$1"
local fstype="$2"

case "$fstype" in
ext2|ext3|ext4)
tune2fs -C 0 -T now "$dev"
$ECHO tune2fs -C 0 -T now -c -1 "$dev"
;;
xfs)
# XFS does not enforce check intervals; nothing to delay
Expand All @@ -146,6 +147,20 @@ function try_delay_checks() {
esac
}

# print the date that $1 was last $3, in a format that date(1) will
# accept, or "Unknown" if we don't know how to find that date.
function try_get_date() {
local dev="$1"
local fstype="$2"
local match="$3"

case "$fstype" in
ext2|ext3|ext4)
dumpe2fs -h "$dev" 2>/dev/null | sed -ne "s/$match:[[:space:]]*//p"
;;
esac
}

# print the date that $1 was last checked, in a format that date(1) will
# accept, or "Unknown" if we don't know how to find that date.
function try_get_check_date() {
Expand All @@ -154,8 +169,7 @@ function try_get_check_date() {

case "$fstype" in
ext2|ext3|ext4)
dumpe2fs -h "$dev" 2>/dev/null | grep 'Last checked:' | \
sed -e 's/Last checked:[[:space:]]*//'
try_get_date "$dev" "$fstype" "Last checked"
;;
*)
# XFS does not save the last-checked date
Expand Down Expand Up @@ -195,25 +209,26 @@ function perform_check() {
# -C 0 is present for cases where the script is run interactively
# (logsave -s strips out the progress bar). ignore the return status
# of this e2fsck, as it doesn't matter.
nice logsave -as "${tmpfile}" e2fsck -p -C 0 "$dev"
$ECHO nice logsave -as "${tmpfile}" e2fsck -p -C 0 "$dev"

# then do the real check; -y is here to give more info on any errors
# that may be present on the FS, in the log file. the snapshot is
# writable, so it shouldn't break anything if e2fsck changes it.
nice logsave -as "${tmpfile}" e2fsck -fy -C 0 "$dev"
$ECHO nice logsave -as "${tmpfile}" e2fsck -fy -C 0 "$dev"
return $?
;;
reiserfs)
echo Yes | nice logsave -as "${tmpfile}" fsck.reiserfs --check "$dev"
# shellcheck disable=SC2216 # nice passes stdin to stdout
echo Yes | $ECHO nice logsave -as "${tmpfile}" fsck.reiserfs --check "$dev"
# apparently can't fail? let's hope not...
return 0
;;
xfs)
nice logsave -as "${tmpfile}" xfs_repair -n "$dev"
$ECHO nice logsave -as "${tmpfile}" xfs_repair -n "$dev"
return $?
;;
jfs)
nice logsave -as "${tmpfile}" fsck.jfs -fn "$dev"
$ECHO nice logsave -as "${tmpfile}" fsck.jfs -fn "$dev"
return $?
;;
*)
Expand All @@ -229,27 +244,30 @@ function check_fs() {
local fstype="$3"
local snapsize="$4"

local tmpfile=`mktemp -t lvcheck.log.XXXXXXXXXX`
local tmpfile
tmpfile=$(mktemp -t lvcheck.log.XXXXXXXXXX)
local errlog="/var/log/lvcheck-${vg}@${lv}"
local snaplvbase="${lv}-lvcheck-temp"
local snaplv="${snaplvbase}-`date +'%Y%m%d'`"
local snaplv
snaplv="${snaplvbase}-$(date +'%Y%m%d')"

# clean up any left-over snapshot LVs
local lvtemp
for lvtemp in /dev/${vg}/${snaplvbase}* ; do
if [ -e "$lvtemp" ] ; then
# Assume the script won't run more than one instance at a time?

log "warning" "Found stale snapshot $lvtemp: attempting to remove."

if ! lvremove -f "${lvtemp##/dev}" ; then
if ! $ECHO lvremove -f "${lvtemp##/dev}" ; then
log "error" "Could not delete stale snapshot $lvtemp"
return 1
fi
fi
done

# and create this one
lvcreate -s -L "${snapsize}M" -n "${snaplv}" "${vg}/${lv}"
$ECHO lvcreate -s -L "${snapsize}M" -n "${snaplv}" "${vg}/${lv}"

if perform_check "/dev/${vg}/${snaplv}" "${fstype}" "${tmpfile}" ; then
log "info" "Background scrubbing of /dev/${vg}/${lv} succeeded."
Expand All @@ -259,7 +277,7 @@ function check_fs() {
try_force_check "/dev/${vg}/${lv}" "$fstype"

if test -n "$EMAIL"; then
mail -s "Fsck of /dev/${vg}/${lv} failed!" $EMAIL < $tmpfile
mail -s "Fsck of /dev/${vg}/${lv} failed!" "$EMAIL" < "$tmpfile"
fi

# save the log file in /var/log in case mail is disabled
Expand All @@ -272,11 +290,21 @@ function check_fs() {
fi

rm -f "$tmpfile"
lvremove -f "${vg}/${snaplv}"
$ECHO lvremove -f "${vg}/${snaplv}"
}

set -e

if [ "$1" = "--dry-run" ]; then
ECHO="echo"
else
ECHO=""
fi

# pull in configuration -- overwrite the defaults above if the file exists
# shellcheck disable=SC1091
[ -r /etc/lvcheck.conf ] && . /etc/lvcheck.conf
[ -r lvcheck.conf ] && . lvcheck.conf

# check whether the machine is on AC power: if not, skip fsck
on_ac_power || exit 0
Expand All @@ -292,44 +320,60 @@ if ! /sbin/dmsetup targets | grep -q snapshot ; then
fi

# parse up lvscan output
lvscan 2>&1 | grep ACTIVE | awk '{print $2;}' | \
while read DEV ; do
# remove the single quotes around the device name
DEV="`echo "$DEV" | tr -d \'`"
if [ -z "$DEVICES" ]; then
if [ -n "$CHECK_DEVICES" ]; then
DEVICES="$CHECK_DEVICES"
else
DEVICES=$(lvscan 2>&1 | sed -ne "/ACTIVE/{s/Original//;s/Snapshot//;s/.*'\\(\\/dev\\/[^']*\\)'.*/\\1/p}")
fi
fi

for DEV in $DEVICES; do
# make sure the device exists
if [ ! -b "$DEV" ]; then
log "warning" "${DEV} doesn't actually exist; skipping"
continue
fi

# get the FS type: blkid prints TYPE="blah"
eval `blkid -s TYPE "$DEV" | cut -d' ' -f2`
TYPE=""
eval "$(blkid -s TYPE "$DEV" | cut -d' ' -f2)"

if [ -z "$TYPE" ]; then
log "warning" "Couldn't determine the filesystem type on ${DEV}; skipping"
continue
fi

# see whether this FS needs any extra checks that might disqualify this device
should_still_check "$DEV" "$TYPE" || continue

# get the last-check time
check_date=`try_get_check_date "$DEV" "$TYPE"`
check_date=$(try_get_check_date "$DEV" "$TYPE")

# if the date is unknown, run fsck every time the script runs. sigh.
if [ "$check_date" != "Unknown" ] ; then
# add $INTERVAL days, and throw away the time portion
check_day=`date --date="$check_date $INTERVAL days" +'%Y%m%d'`
check_day=$(date --date="$check_date $INTERVAL days" +'%Y%m%d')

# get today's date, and skip the check if it's not within the interval
today=`date +'%Y%m%d'`
[ $check_day -gt $today ] && continue
today=$(date +'%Y%m%d')
[ "$check_day" -gt "$today" ] && continue
fi

# get the volume group and logical volume names
VG="`lvs --noheadings -o vg_name "$DEV" | tr -d ' '`"
LV="`lvs --noheadings -o lv_name "$DEV" | tr -d ' '`"
VG="$(lvs --noheadings -o vg_name "$DEV" | tr -d ' ')"
LV="$(lvs --noheadings -o lv_name "$DEV" | tr -d ' ')"

# get the free space and LV size (in megs), guess at the snapshot
# size, and see how much the admin will let us use (keeping MINFREE
# available)
SPACE="`lvs --noheadings --units M --nosuffix -o vg_free "$DEV" | tr -d ' '`"
SIZE="`lvs --noheadings --units M --nosuffix -o lv_size "$DEV" | tr -d ' '`"
SNAPSIZE="`expr "${SIZE%%.*}" / 500`"
AVAIL="`expr "${SPACE%%.*}" - "$MINFREE"`"
SPACE="$(lvs --noheadings --units M --nosuffix -o vg_free "$DEV" | tr -d ' ')"
SIZE="$(lvs --noheadings --units M --nosuffix -o lv_size "$DEV" | tr -d ' ')"
SNAPSIZE="$((${SIZE%%.*} / 500))"
AVAIL="$((${SPACE%%.*} - MINFREE))"

# if we don't even have MINSNAP space available, skip the LV
if [ "$MINSNAP" -gt "$AVAIL" -o "$AVAIL" -le 0 ] ; then
if [ "$MINSNAP" -gt "$AVAIL" ] || [ "$AVAIL" -le 0 ] ; then
log "warning" "Not enough free space on volume group for ${DEV}; skipping"
continue
fi
Expand Down
5 changes: 5 additions & 0 deletions lvcheck.conf
View file
Original file line number Diff line number Diff line change
Expand Up @@ -44,3 +44,8 @@

#MINFREE=0

# CHECK_DEVICES
# A list of LVs to check rather than trying to check all
# LVs

#CHECK_DEVICES=

0 comments on commit f8f5b04

Please sign in to comment.