shaba/lvm2
1
0
Fork 0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-22 17:35:59 +03:00
Code Releases Activity
7825d914e3
lvm2/test/shell/metadata-bad-text.sh

323 lines
6.2 KiB
Bash
Raw Normal View History

tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
#!/usr/bin/env bash
# Copyright (C) 2008-2013,2018 Red Hat, Inc. All rights reserved.
#
# This copyrighted material is made available to anyone wishing to use,
# modify, copy, or redistribute it subject to the terms and conditions
# of the GNU General Public License v.2.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software Foundation,
# Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
tests: skip more tests for lvmpolld pass These test do not test polling, so skip them for lvmpolld pass.
2021-03-26 18:05:28 +03:00
SKIP_WITH_LVMPOLLD=1
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
RUNDIR="/run"
test -d "$RUNDIR" || RUNDIR="/var/run"
PVS_ONLINE_DIR="$RUNDIR/lvm/pvs_online"
VGS_ONLINE_DIR="$RUNDIR/lvm/vgs_online"
_clear_online_files() {
# wait till udev is finished
aux udev_wait
rm -f "$PVS_ONLINE_DIR"/*
rm -f "$VGS_ONLINE_DIR"/*
}
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
. lib/inittest
aux prepare_devs 3
get_devs
tests: use aux clear_dev Use wrapper over 'blkdiscard -z'
2023-04-21 14:16:22 +03:00
aux clear_devs "$dev1" "$dev2" "$dev3"
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
vgcreate $SHARED $vg "$dev1" "$dev2" "$dev3"
pvs
dd if="$dev1" of=meta1 bs=4k count=2
sed 's/flags =/flagx =/' meta1 > meta1.bad
dd if=meta1.bad of="$dev1"
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
grep "Checksum error" out
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
# bad metadata in one mda doesn't prevent using
# the VG since other mdas are fine and usable
lvcreate -l1 $vg
vgck --updatemetadata $vg
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
not grep "Checksum error" out
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
vgchange -an $vg
vgremove -ff $vg
#
# Same test as above, but corrupt metadata text
# on two of the three PVs, leaving one good
# copy of the metadata.
#
tests: use aux clear_dev Use wrapper over 'blkdiscard -z'
2023-04-21 14:16:22 +03:00
aux clear_devs "$dev1" "$dev2" "$dev3"
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
vgcreate $SHARED $vg "$dev1" "$dev2" "$dev3"
pvs
dd if="$dev1" of=meta1 bs=4k count=2
dd if="$dev2" of=meta2 bs=4k count=2
sed 's/READ/RRRR/' meta1 > meta1.bad
sed 's/seqno =/sss =/' meta2 > meta2.bad
dd if=meta1.bad of="$dev1"
dd if=meta2.bad of="$dev2"
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
# FIXME: find a better test than looking for a specific message
grep "Checksum error" out > out2
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
grep "$dev1" out2
grep "$dev2" out2
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
# bad metadata in one mda doesn't prevent using
# the VG since other mdas are fine
lvcreate -l1 $vg
vgck --updatemetadata $vg
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
not grep "Checksum error" out
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
vgchange -an $vg
vgremove -ff $vg
#
# Three PVs where two have one mda, and the third
# has two mdas. The first mda is corrupted on all
# thee PVs, but the second mda on the third PV
# makes the VG usable.
#
tests: use aux clear_dev Use wrapper over 'blkdiscard -z'
2023-04-21 14:16:22 +03:00
aux clear_devs "$dev1" "$dev2" "$dev3"
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
pvcreate "$dev1"
pvcreate "$dev2"
pvcreate --pvmetadatacopies 2 "$dev3"
vgcreate $SHARED $vg "$dev1" "$dev2" "$dev3"
pvs
dd if="$dev1" of=meta1 bs=4k count=2
dd if="$dev2" of=meta2 bs=4k count=2
dd if="$dev3" of=meta3 bs=4k count=2
sed 's/READ/RRRR/' meta1 > meta1.bad
sed 's/seqno =/sss =/' meta2 > meta2.bad
sed 's/id =/id/' meta3 > meta3.bad
dd if=meta1.bad of="$dev1"
dd if=meta2.bad of="$dev2"
dd if=meta3.bad of="$dev3"
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
grep "Checksum error" out > out2
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
grep "$dev1" out2
grep "$dev2" out2
grep "$dev3" out2
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
# bad metadata in some mdas doesn't prevent using
# the VG if there's a good mda found
lvcreate -l1 $vg
vgck --updatemetadata $vg
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
not grep "Checksum error" out
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
vgchange -an $vg
vgremove -ff $vg
#
# Test that vgck --updatemetadata will update old metadata
# and repair bad metadata text at the same time from different
# devices.
#
tests: use aux clear_dev Use wrapper over 'blkdiscard -z'
2023-04-21 14:16:22 +03:00
aux clear_devs "$dev1" "$dev2" "$dev3"
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
pvcreate "$dev1"
pvcreate "$dev2"
pvcreate --pvmetadatacopies 2 "$dev3"
vgcreate $SHARED $vg "$dev1" "$dev2" "$dev3"
# Put bad metadata onto dev1
dd if="$dev1" of=meta1 bs=4k count=2
sed 's/READ/RRRR/' meta1 > meta1.bad
dd if=meta1.bad of="$dev1"
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
grep "Checksum error" out > out2
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
grep "$dev1" out2
# We can still use the VG with other available
# mdas, skipping the bad mda.
lvcreate -n $lv1 -l1 -an $vg "$dev1"
lvcreate -n $lv2 -l1 -an $vg "$dev1"
# Put old metadata onto dev2 by updating
# the VG while dev2 is disabled.
aux disable_dev "$dev2"
pvs
pvs "$dev1"
not pvs "$dev2"
pvs "$dev3"
lvs $vg/$lv1
lvs $vg/$lv2
lvremove $vg/$lv2
aux enable_dev "$dev2"
# Both old and bad metadata are reported.
pvs 2>&1 | tee out
grep "ignoring metadata seqno" out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
grep "Checksum error" out
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
lvs $vg/$lv1
not lvs $vg/$lv2
# fixes the bad metadata on dev1, and
# fixes the old metadata on dev2.
vgck --updatemetadata $vg
pvs 2>&1 | tee out
not grep "ignoring metadata seqno" out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
not grep "Checksum error" out
tests: add metadata-bad-text.sh
2019-05-24 23:58:05 +03:00
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
lvs $vg/$lv1
vgchange -an $vg
vgremove -ff $vg
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
#
# Test pvscan activation with bad PVs
#
tests: metadata-bad-text.sh with lvmlockd skip part of test that doesn't apply
2019-09-05 23:27:52 +03:00
# autoactivation not done on shared VGs
if test -n "$LVM_TEST_LVMLOCKD"; then
exit 0
fi
tests: use aux clear_dev Use wrapper over 'blkdiscard -z'
2023-04-21 14:16:22 +03:00
aux clear_devs "$dev1" "$dev2" "$dev3"
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
vgcreate $SHARED $vg "$dev1" "$dev2" "$dev3"
tests: reduce shellcheck warns Reduce warnings in tests.
2023-02-09 18:05:40 +03:00
PVID1=$(pvs "$dev1" --noheading -o uuid | tr -d - | awk '{print $1}')
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
echo $PVID1
tests: reduce shellcheck warns Reduce warnings in tests.
2023-02-09 18:05:40 +03:00
PVID2=$(pvs "$dev2" --noheading -o uuid | tr -d - | awk '{print $1}')
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
echo $PVID2
tests: reduce shellcheck warns Reduce warnings in tests.
2023-02-09 18:05:40 +03:00
PVID3=$(pvs "$dev3" --noheading -o uuid | tr -d - | awk '{print $1}')
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
echo $PVID3
pvs
dd if="$dev1" of=meta1 bs=4k count=2
dd if="$dev2" of=meta2 bs=4k count=2
sed 's/READ/RRRR/' meta1 > meta1.bad
sed 's/seqno =/sss =/' meta2 > meta2.bad
dd if=meta1.bad of="$dev1"
dd if=meta2.bad of="$dev2"
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
grep "Checksum error" out > out2
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
grep "$dev1" out2
grep "$dev2" out2
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
# bad metadata in one mda doesn't prevent using
# the VG since other mdas are fine
lvcreate -l1 -n $lv1 $vg
vgchange -an $vg
_clear_online_files
tests: explicitely use event_activation=1 Some test do expect event_activation to be set. So add explicit configuring of this setting in tests, but also add new default which kind of does it globally as it's expected default (yet our testing rpms might be create with disabled event_activation) By adding this to each test individually - it's now easy to locate such tests...
2024-04-11 17:21:11 +03:00
aux lvmconf "global/event_activation = 1"
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
# pvscan of one dev with bad metadata will result
pvscan: rework to improve PVs without metadata Restructure the pvscan code, and add new temporary files that list pvids in a VG, used for processing PVs that have no metadata. The new temp files, in /run/lvm/pvs_lookup/<vgname>, allow a proper pvscan --cache to be done on PVs that have no metadata. pvscan --cache <dev> is only supposed to read <dev>, but when <dev> has no metadata, this had not been possible. The command had to fall back to scanning all devices to read all VG metadata to get the list of all PVIDs needed to check for a complete VG. Now, the temp file can be used in place of reading metadata from all PVs on the system.
2020-10-15 22:11:08 +03:00
# in the pvid online file being created but the
# VG will not be known.
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
pvscan --cache -aay "$dev1"
ls "$RUNDIR/lvm/pvs_online/$PVID1"
pvscan: rework to improve PVs without metadata Restructure the pvscan code, and add new temporary files that list pvids in a VG, used for processing PVs that have no metadata. The new temp files, in /run/lvm/pvs_lookup/<vgname>, allow a proper pvscan --cache to be done on PVs that have no metadata. pvscan --cache <dev> is only supposed to read <dev>, but when <dev> has no metadata, this had not been possible. The command had to fall back to scanning all devices to read all VG metadata to get the list of all PVIDs needed to check for a complete VG. Now, the temp file can be used in place of reading metadata from all PVs on the system.
2020-10-15 22:11:08 +03:00
not ls "$RUNDIR/lvm/pvs_online/$PVID2"
not ls "$RUNDIR/lvm/pvs_online/$PVID3"
not ls "$RUNDIR/lvm/vgs_online/$vg"
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
_clear_online_files
# scan the one pv with good metadata, does not scan any others
pvscan --cache -aay "$dev3"
not ls "$RUNDIR/lvm/pvs_online/$PVID1"
not ls "$RUNDIR/lvm/pvs_online/$PVID2"
ls "$RUNDIR/lvm/pvs_online/$PVID3"
not ls "$RUNDIR/lvm/vgs_online/$vg"
vgck --updatemetadata $vg
pvs 2>&1 | tee out
handle bad metadata text in vg_read path Corrupt metadata text (with good mda header) was being handled in the label_scan phase, but not in the vg_read phase. This was sufficient because metadata areas would always be read and checksummed during label_scan (metadata parsing was skipped previously as an optimization.) This changed with the optimization in commit 61a6f9905e87e650f0bddae83fec6923bb100a57 "metadata: optimize reading metadata copies in scan" Now, some metadata areas will not be read and checksummed at all during the label_scan phase, only during the vg_read phase. This means that bad metadata text may first be detected in the vg_read phase. So, add equivalent bad metadata handling to the vg_read path to match the label_scan path.
2021-09-28 22:58:03 +03:00
not grep "Checksum error" out
pvscan: avoid full scan for activation When an online PV completed a VG, the standard activation functions were used to activate the VG. These functions use a full scan of all devs. When many pvscans are run during startup and need to activate many VGs, scanning all devs from all the pvscans can take a long time. Optimize VG activation in pvscan to scan only the devs in the VG being activated. This makes use of the online file info that was used to determine the VG was complete. The downside of this approach is that pvscan activation will not detect duplicate PVs and block activation, where a normal activation command (which scans all devices) would.
2019-08-27 01:07:18 +03:00
pvs "$dev1"
pvs "$dev2"
pvs "$dev3"
vgchange -an $vg
vgremove -ff $vg
Copy Permalink