RAID-Z expansion feature

This feature allows disks to be added one at a time to a RAID-Z group,
expanding its capacity incrementally.  This feature is especially useful
for small pools (typically with only one RAID-Z group), where there
isn't sufficient hardware to add capacity by adding a whole new RAID-Z
group (typically doubling the number of disks).

== Initiating expansion ==

A new device (disk) can be attached to an existing RAIDZ vdev, by
running `zpool attach POOL raidzP-N NEW_DEVICE`, e.g. `zpool attach tank
raidz2-0 sda`.  The new device will become part of the RAIDZ group.  A
"raidz expansion" will be initiated, and the new device will contribute
additional space to the RAIDZ group once the expansion completes.

The `feature@raidz_expansion` on-disk feature flag must be `enabled` to
initiate an expansion, and it remains `active` for the life of the pool.
In other words, pools with expanded RAIDZ vdevs can not be imported by
older releases of the ZFS software.

== During expansion ==

The expansion entails reading all allocated space from existing disks in
the RAIDZ group, and rewriting it to the new disks in the RAIDZ group
(including the newly added device).

The expansion progress can be monitored with `zpool status`.

Data redundancy is maintained during (and after) the expansion.  If a
disk fails while the expansion is in progress, the expansion pauses
until the health of the RAIDZ vdev is restored (e.g. by replacing the
failed disk and waiting for reconstruction to complete).

The pool remains accessible during expansion.  Following a reboot or
export/import, the expansion resumes where it left off.

== After expansion ==

When the expansion completes, the additional space is available for use,
and is reflected in the `available` zfs property (as seen in `zfs list`,
`df`, etc).

Expansion does not change the number of failures that can be tolerated
without data loss (e.g. a RAIDZ2 is still a RAIDZ2 even after
expansion).

A RAIDZ vdev can be expanded multiple times.

After the expansion completes, old blocks remain with their old
data-to-parity ratio (e.g. 5-wide RAIDZ2, has 3 data to 2 parity), but
distributed among the larger set of disks.  New blocks will be written
with the new data-to-parity ratio (e.g. a 5-wide RAIDZ2 which has been
expanded once to 6-wide, has 4 data to 2 parity).  However, the RAIDZ
vdev's "assumed parity ratio" does not change, so slightly less space
than is expected may be reported for newly-written blocks, according to
`zfs list`, `df`, `ls -s`, and similar tools.

Sponsored-by: The FreeBSD Foundation
Sponsored-by: iXsystems, Inc.
Sponsored-by: vStack
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Authored-by: Matthew Ahrens <mahrens@delphix.com>
Contributions-by: Fedor Uporov <fuporov.vstack@gmail.com>
Contributions-by: Stuart Maybee <stuart.maybee@comcast.net>
Contributions-by: Thorsten Behrens <tbehrens@outlook.com>
Contributions-by: Fmstrat <nospam@nowsci.com>
Contributions-by: Don Brady <dev.fs.zfs@gmail.com>
Signed-off-by: Don Brady <dev.fs.zfs@gmail.com>
Closes #15022
This commit is contained in:
Don Brady
2023-11-08 11:19:41 -07:00
committed by GitHub
parent 9198de8f10
commit 5caeef02fa
62 changed files with 5740 additions and 876 deletions
+5 -1
View File
@@ -769,8 +769,12 @@ tests = ['redacted_compressed', 'redacted_contents', 'redacted_deleted',
tags = ['functional', 'redacted_send']
[tests/functional/raidz]
tests = ['raidz_001_neg', 'raidz_002_pos', 'raidz_003_pos', 'raidz_004_pos']
tests = ['raidz_001_neg', 'raidz_002_pos', 'raidz_expand_001_pos',
'raidz_expand_002_pos', 'raidz_expand_003_neg', 'raidz_expand_003_pos',
'raidz_expand_004_pos', 'raidz_expand_005_pos', 'raidz_expand_006_neg',
'raidz_expand_007_neg']
tags = ['functional', 'raidz']
timeout = 1200
[tests/functional/redundancy]
tests = ['redundancy_draid', 'redundancy_draid1', 'redundancy_draid2',
+3
View File
@@ -34,6 +34,7 @@ DEADMAN_SYNCTIME_MS deadman.synctime_ms zfs_deadman_synctime_ms
DEADMAN_ZIOTIME_MS deadman.ziotime_ms zfs_deadman_ziotime_ms
DISABLE_IVSET_GUID_CHECK disable_ivset_guid_check zfs_disable_ivset_guid_check
DMU_OFFSET_NEXT_SYNC dmu_offset_next_sync zfs_dmu_offset_next_sync
EMBEDDED_SLOG_MIN_MS embedded_slog_min_ms zfs_embedded_slog_min_ms
INITIALIZE_CHUNK_SIZE initialize_chunk_size zfs_initialize_chunk_size
INITIALIZE_VALUE initialize_value zfs_initialize_value
KEEP_LOG_SPACEMAPS_AT_EXPORT keep_log_spacemaps_at_export zfs_keep_log_spacemaps_at_export
@@ -62,6 +63,7 @@ MULTIHOST_IMPORT_INTERVALS multihost.import_intervals zfs_multihost_import_inter
MULTIHOST_INTERVAL multihost.interval zfs_multihost_interval
OVERRIDE_ESTIMATE_RECORDSIZE send.override_estimate_recordsize zfs_override_estimate_recordsize
PREFETCH_DISABLE prefetch.disable zfs_prefetch_disable
RAIDZ_EXPAND_MAX_REFLOW_BYTES vdev.expand_max_reflow_bytes raidz_expand_max_reflow_bytes
REBUILD_SCRUB_ENABLED rebuild_scrub_enabled zfs_rebuild_scrub_enabled
REMOVAL_SUSPEND_PROGRESS removal_suspend_progress zfs_removal_suspend_progress
REMOVE_MAX_SEGMENT remove_max_segment zfs_remove_max_segment
@@ -69,6 +71,7 @@ RESILVER_MIN_TIME_MS resilver_min_time_ms zfs_resilver_min_time_ms
SCAN_LEGACY scan_legacy zfs_scan_legacy
SCAN_SUSPEND_PROGRESS scan_suspend_progress zfs_scan_suspend_progress
SCAN_VDEV_LIMIT scan_vdev_limit zfs_scan_vdev_limit
SCRUB_AFTER_EXPAND scrub_after_expand zfs_scrub_after_expand
SEND_HOLES_WITHOUT_BIRTH_TIME send_holes_without_birth_time send_holes_without_birth_time
SLOW_IO_EVENTS_PER_SECOND slow_io_events_per_second zfs_slow_io_events_per_second
SPA_ASIZE_INFLATION spa.asize_inflation spa_asize_inflation
+8 -2
View File
@@ -1668,8 +1668,14 @@ nobase_dist_datadir_zfs_tests_tests_SCRIPTS += \
functional/raidz/cleanup.ksh \
functional/raidz/raidz_001_neg.ksh \
functional/raidz/raidz_002_pos.ksh \
functional/raidz/raidz_003_pos.ksh \
functional/raidz/raidz_004_pos.ksh \
functional/raidz/raidz_expand_001_pos.ksh \
functional/raidz/raidz_expand_002_pos.ksh \
functional/raidz/raidz_expand_003_neg.ksh \
functional/raidz/raidz_expand_003_pos.ksh \
functional/raidz/raidz_expand_004_pos.ksh \
functional/raidz/raidz_expand_005_pos.ksh \
functional/raidz/raidz_expand_006_neg.ksh \
functional/raidz/raidz_expand_007_neg.ksh \
functional/raidz/setup.ksh \
functional/redacted_send/cleanup.ksh \
functional/redacted_send/redacted_compressed.ksh \
@@ -106,5 +106,6 @@ if is_linux || is_freebsd; then
"feature@blake3"
"feature@block_cloning"
"feature@vdev_zaps_v2"
"feature@raidz_expansion"
)
fi
@@ -23,19 +23,39 @@
#
# Copyright (c) 2016 by Gvozden Neskovic. All rights reserved.
# Use is subject to license terms.
# Copyright (c) 2020 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# Call the raidz_test tool with -S to test all supported raidz
# implementations. This options will test several raidz block geometries
# Call the raidz_test tool with sweep to test all supported raidz
# implementations. This will test several raidz block geometries
# and several zio parameters that affect raidz block layout. Data
# reconstruction performs all combinations of failed disks. Wall time
# is set to 5min, but actual runtime might be longer.
# reconstruction performs all combinations of failed disks. Wall
# time is set to 5 min, but actual runtime might be longer.
#
log_must raidz_test -S -t 300
log_pass "raidz_test parameter sweep test succeeded."
case $((RANDOM % 3)) in
0)
# Basic sweep test
log_must raidz_test -S -t 300
log_pass "raidz_test parameter sweep test succeeded."
;;
1)
# Using expanded raidz map to test all supported raidz
# implementations with expanded map and default reflow offset.
log_must raidz_test -S -e -t 300
log_pass "raidz_test sweep test with expanded map succeeded."
;;
2)
# Using expanded raidz map ('-e') to test all supported raidz
# implementations with expanded map and zero reflow offset.
log_must raidz_test -S -e -r 0 -t 300
log_pass "raidz_test sweep test with expanded map succeeded."
;;
*)
# avoid shellcheck SC2249
;;
esac
@@ -1,41 +0,0 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or https://opensource.org/licenses/CDDL-1.0.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2020 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# Call the raidz_test tool with -S and -e to test all supported raidz
# implementations with expanded map and default reflow offset.
# This options will test several raidz block geometries and several zio
# parameters that affect raidz block layout. Data reconstruction performs
# all combinations of failed disks. Wall time is set to 5min, but actual
# runtime might be longer.
#
log_must raidz_test -S -e -t 60
log_pass "raidz_test parameter sweep test with expanded map succeeded."
@@ -1,41 +0,0 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or https://opensource.org/licenses/CDDL-1.0.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2020 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# Call the raidz_test tool with -S and -e to test all supported raidz
# implementations with expanded map and zero reflow offset.
# This options will test several raidz block geometries and several zio
# parameters that affect raidz block layout. Data reconstruction performs
# all combinations of failed disks. Wall time is set to 5min, but actual
# runtime might be longer.
#
log_must raidz_test -S -e -r 0 -t 60
log_pass "raidz_test parameter sweep test with expanded map succeeded."
@@ -0,0 +1,215 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2020 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# 'zpool attach poolname raidz ...' should attach new device to the pool.
#
# STRATEGY:
# 1. Create block device files for the test raidz pool
# 2. For each parity value [1..3]
# - create raidz pool
# - fill it with some directories/files
# - attach device to the raidz pool
# - verify that device attached and the raidz pool size increase
# - verify resilver by replacing parity devices
# - verify resilver by replacing data devices
# - verify scrub by zeroing parity devices
# - verify scrub by zeroing data devices
# - verify the raidz pool
# - destroy the raidz pool
typeset -r devs=6
typeset -r dev_size_mb=128
typeset -a disks
prefetch_disable=$(get_tunable PREFETCH_DISABLE)
function cleanup
{
log_pos zpool status $TESTPOOL
poolexists "$TESTPOOL" && log_must_busy zpool destroy "$TESTPOOL"
for i in {0..$devs}; do
log_must rm -f "$TEST_BASE_DIR/dev-$i"
done
log_must set_tunable32 PREFETCH_DISABLE $prefetch_disable
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES 0
}
function wait_expand_paused
{
oldcopied='0'
newcopied='1'
while [[ $oldcopied != $newcopied ]]; do
oldcopied=$newcopied
sleep 2
newcopied=$(zpool status $TESTPOOL | \
grep 'copied out of' | \
awk '{print $1}')
log_note "newcopied=$newcopied"
done
log_note "paused at $newcopied"
}
function test_resilver # <pool> <parity> <dir>
{
typeset pool=$1
typeset nparity=$2
typeset dir=$3
for (( i=0; i<$nparity; i=i+1 )); do
log_must zpool offline $pool $dir/dev-$i
done
log_must zpool export $pool
for (( i=0; i<$nparity; i=i+1 )); do
log_must zpool labelclear -f $dir/dev-$i
done
log_must zpool import -o cachefile=none -d $dir $pool
for (( i=0; i<$nparity; i=i+1 )); do
log_must zpool replace -f $pool $dir/dev-$i
done
log_must zpool wait -t resilver $pool
log_must check_pool_status $pool "errors" "No known data errors"
log_must zpool clear $pool
for (( i=$nparity; i<$nparity*2; i=i+1 )); do
log_must zpool offline $pool $dir/dev-$i
done
log_must zpool export $pool
for (( i=$nparity; i<$nparity*2; i=i+1 )); do
log_must zpool labelclear -f $dir/dev-$i
done
log_must zpool import -o cachefile=none -d $dir $pool
for (( i=$nparity; i<$nparity*2; i=i+1 )); do
log_must zpool replace -f $pool $dir/dev-$i
done
log_must zpool wait -t resilver $pool
log_must check_pool_status $pool "errors" "No known data errors"
log_must zpool clear $pool
}
function test_scrub # <pool> <parity> <dir>
{
typeset pool=$1
typeset nparity=$2
typeset dir=$3
typeset combrec=$4
reflow_size=$(get_pool_prop allocated $pool)
randbyte=$(( ((RANDOM<<15) + RANDOM) % $reflow_size ))
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES $randbyte
log_must zpool attach $TESTPOOL ${raid}-0 $dir/dev-$devs
wait_expand_paused
log_must zpool export $pool
# zero out parity disks
for (( i=0; i<$nparity; i=i+1 )); do
dd conv=notrunc if=/dev/zero of=$dir/dev-$i \
bs=1M seek=4 count=$(($dev_size_mb-4))
done
log_must zpool import -o cachefile=none -d $dir $pool
log_must zpool scrub -w $pool
log_must zpool clear $pool
log_must zpool export $pool
# zero out parity count worth of data disks
for (( i=$nparity; i<$nparity*2; i=i+1 )); do
dd conv=notrunc if=/dev/zero of=$dir/dev-$i \
bs=1M seek=4 count=$(($dev_size_mb-4))
done
log_must zpool import -o cachefile=none -d $dir $pool
log_must zpool scrub -w $pool
log_must check_pool_status $pool "errors" "No known data errors"
log_must zpool clear $pool
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES 0
log_must zpool wait -t raidz_expand $TESTPOOL
}
log_onexit cleanup
log_must set_tunable32 PREFETCH_DISABLE 1
# Disk files which will be used by pool
for i in {0..$(($devs - 1))}; do
device=$TEST_BASE_DIR/dev-$i
log_must truncate -s ${dev_size_mb}M $device
disks[${#disks[*]}+1]=$device
done
# Disk file which will be attached
log_must truncate -s 512M $TEST_BASE_DIR/dev-$devs
nparity=$((RANDOM%(3) + 1))
raid=raidz$nparity
dir=$TEST_BASE_DIR
log_must zpool create -f -o cachefile=none $TESTPOOL $raid ${disks[@]}
log_must zfs set primarycache=metadata $TESTPOOL
log_must zfs create $TESTPOOL/fs
log_must fill_fs /$TESTPOOL/fs 1 512 100 1024 R
log_must zfs create -o compress=on $TESTPOOL/fs2
log_must fill_fs /$TESTPOOL/fs2 1 512 100 1024 R
log_must zfs create -o compress=on -o recordsize=8k $TESTPOOL/fs3
log_must fill_fs /$TESTPOOL/fs3 1 512 100 1024 R
log_must check_pool_status $TESTPOOL "errors" "No known data errors"
test_scrub $TESTPOOL $nparity $dir
test_resilver $TESTPOOL $nparity $dir
zpool destroy "$TESTPOOL"
log_pass "raidz expansion test succeeded."
@@ -0,0 +1,115 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2020 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# 'zpool attach poolname raidz ...' should attach new devive to the pool.
#
# STRATEGY:
# 1. Create block device files for the test raidz pool
# 2. For each parity value [1..3]
# - create raidz pool with minimum block device files required
# - for each free test block device
# - attach to the pool
# - verify the raidz pool
# - destroy the raidz pool
typeset -r devs=6
typeset -r dev_size_mb=512
typeset -a disks
prefetch_disable=$(get_tunable PREFETCH_DISABLE)
function cleanup
{
poolexists "$TESTPOOL" && log_must_busy zpool destroy "$TESTPOOL"
for i in {0..$devs}; do
log_must rm -f "$TEST_BASE_DIR/dev-$i"
done
log_must set_tunable32 PREFETCH_DISABLE $prefetch_disable
}
log_onexit cleanup
log_must set_tunable32 PREFETCH_DISABLE 1
# Disk files which will be used by pool
for i in {0..$(($devs))}; do
device=$TEST_BASE_DIR/dev-$i
log_must truncate -s ${dev_size_mb}M $device
disks[${#disks[*]}+1]=$device
done
nparity=$((RANDOM%(3) + 1))
raid=raidz$nparity
dir=$TEST_BASE_DIR
pool=$TESTPOOL
opts="-o cachefile=none"
log_must zpool create -f $opts $pool $raid ${disks[1..$(($nparity+1))]}
log_must zfs set primarycache=metadata $pool
log_must zfs create $pool/fs
log_must fill_fs /$pool/fs 1 512 100 1024 R
log_must zfs create -o compress=on $pool/fs2
log_must fill_fs /$pool/fs2 1 512 100 1024 R
log_must zfs create -o compress=on -o recordsize=8k $pool/fs3
log_must fill_fs /$pool/fs3 1 512 100 1024 R
typeset pool_size=$(get_pool_prop size $pool)
for disk in ${disks[$(($nparity+2))..$devs]}; do
log_must dd if=/dev/urandom of=/${pool}/FILE-$RANDOM bs=1M \
count=64
log_must zpool attach -w $pool ${raid}-0 $disk
# Wait some time for pool size increase
sleep 5
# Confirm that disk was attached to the pool
log_must zpool get -H path $TESTPOOL $disk
typeset expand_size=$(get_pool_prop size $pool)
if [[ "$expand_size" -le "$pool_size" ]]; then
log_fail "pool $pool not expanded"
fi
verify_pool $pool
pool_size=$expand_size
done
zpool destroy "$pool"
log_pass "raidz expansion test succeeded."
@@ -0,0 +1,102 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2021 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# 'zpool attach poolname raidz ...' should reject device attach if pool
# is in checkpointed state. If checkpoint creation requested on
# expanding pool, the request should be rejected.
#
# STRATEGY:
# 1. Create block device files for the test raidz pool.
# 2. Create pool and checkpoint it.
# 3. Try to expand raidz, ensure that request rejected.
# 4. Recreate the pool.
# 5. Apply raidz expansion.
# 6. Ensure, that checkpoint cannot be created.
typeset -r devs=6
typeset -r dev_size_mb=512
typeset -a disks
prefetch_disable=$(get_tunable PREFETCH_DISABLE)
function cleanup
{
poolexists "$TESTPOOL" && log_must_busy zpool destroy "$TESTPOOL"
for i in {0..$devs}; do
log_must rm -f "$TEST_BASE_DIR/dev-$i"
done
log_must set_tunable32 PREFETCH_DISABLE $prefetch_disable
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES 0
}
log_onexit cleanup
log_must set_tunable32 PREFETCH_DISABLE 1
# Disk files which will be used by pool
for i in {0..$(($devs))}; do
device=$TEST_BASE_DIR/dev-$i
log_must truncate -s ${dev_size_mb}M $device
disks[${#disks[*]}+1]=$device
done
nparity=1
raid=raidz$nparity
pool=$TESTPOOL
opts="-o cachefile=none"
# case 1: checkpoint exist, try to expand
log_must zpool create -f $opts $pool $raid ${disks[1..$(($devs-1))]}
log_must zfs set primarycache=metadata $pool
log_must zpool checkpoint $pool
log_mustnot zpool attach $pool ${raid}-0 ${disks[$devs]}
log_must zpool destroy $pool
#
# case 2: expansion in progress, try to checkpoint
#
# Sets pause point at 25% of allocated space so that we know an
# expansion is still in progress when we attempt the checkpoint
#
log_must zpool create -f $opts $pool $raid ${disks[1..$(($devs-1))]}
log_must zfs set primarycache=metadata $pool
log_must zfs create $pool/fs
log_must fill_fs /$pool/fs 1 512 100 1024 R
allocated=$(zpool list -Hp -o allocated $pool)
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES $((allocated / 4))
log_must zpool attach $pool ${raid}-0 ${disks[$devs]}
log_mustnot zpool checkpoint $pool
log_must zpool destroy $pool
log_pass "raidz expansion test succeeded."
@@ -0,0 +1,141 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2021 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
. $STF_SUITE/tests/functional/cli_root/zfs_wait/zfs_wait.kshlib
#
# DESCRIPTION:
# Check raidz expansion is able to work correctly under i/o load.
#
# STRATEGY:
# 1. Create block device files for the test raidz pool
# 2. For each parity value [1..3]
# - create raidz pool with minimum block device files required
# - create couple of datasets with different recordsize and fill it
# - set a max reflow value near pool capacity
# - wait for reflow to reach this max
# - verify pool
# - set reflow bytes to max value to complete the expansion
typeset -r devs=10
typeset -r dev_size_mb=256
typeset -a disks
embedded_slog_min_ms=$(get_tunable EMBEDDED_SLOG_MIN_MS)
function cleanup
{
poolexists "$TESTPOOL" && zpool status -v "$TESTPOOL"
poolexists "$TESTPOOL" && log_must_busy zpool destroy "$TESTPOOL"
for i in {0..$devs}; do
log_must rm -f "$TEST_BASE_DIR/dev-$i"
done
log_must set_tunable32 EMBEDDED_SLOG_MIN_MS $embedded_slog_min_ms
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES 0
}
function wait_expand_paused
{
oldcopied='0'
newcopied='1'
# wait until reflow copied value stops changing
while [[ $oldcopied != $newcopied ]]; do
oldcopied=$newcopied
sleep 1
newcopied=$(zpool status $TESTPOOL | \
grep 'copied out of' | \
awk '{print $1}')
done
}
log_onexit cleanup
log_must set_tunable32 EMBEDDED_SLOG_MIN_MS 99999
# Disk files which will be used by pool
for i in {0..$(($devs))}; do
device=$TEST_BASE_DIR/dev-$i
log_must truncate -s ${dev_size_mb}M $device
disks[${#disks[*]}+1]=$device
done
nparity=$((RANDOM%(3) + 1))
raid=raidz$nparity
pool=$TESTPOOL
opts="-o cachefile=none"
log_must zpool create -f $opts $pool $raid ${disks[1..$(($nparity+1))]}
log_must zfs create -o recordsize=8k $pool/fs
log_must fill_fs /$pool/fs 1 256 100 1024 R
log_must zfs create -o recordsize=128k $pool/fs2
log_must fill_fs /$pool/fs2 1 256 100 1024 R
for disk in ${disks[$(($nparity+2))..$devs]}; do
log_must mkfile -n 400m /$pool/fs/file
log_bkgrnd randwritecomp /$pool/fs/file 250
pid0=$!
# start some random writes in the background during expansion
log_must mkfile -n 400m /$pool/fs2/file2
log_bkgrnd randwritecomp /$pool/fs2/file2 250
pid1=$!
sleep 10
# Pause at half total bytes to be copied for expansion
reflow_size=$(get_pool_prop allocated $pool)
log_note need to reflow $reflow_size bytes
pause=$((reflow_size/2))
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES $pause
log_must zpool attach $pool ${raid}-0 $disk
wait_expand_paused
kill_if_running $pid0
kill_if_running $pid1
log_must zpool scrub -w $pool
log_must check_pool_status $pool "errors" "No known data errors"
log_must check_pool_status $pool "scan" "with 0 errors"
log_must check_pool_status $pool "scan" "repaired 0B"
# Set pause past largest possible value for this pool
pause=$((devs*dev_size_mb*1024*1024))
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES $pause
log_must zpool wait -t raidz_expand $pool
done
log_must zpool destroy "$pool"
log_pass "raidz expansion test succeeded."
@@ -0,0 +1,121 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2021 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# Check device replacement during raidz expansion.
#
# STRATEGY:
# 1. Create block device files for the test raidz pool
# 2. For each parity value [1..3]
# - create raidz pool with minimum block device files required
# - create couple of datasets with different recordsize and fill it
# - attach new device to the pool
# - offline and zero vdevs allowed by parity
# - wait some time and start offlined vdevs replacement
# - wait replacement completion and verify pool status
typeset -r devs=10
typeset -r dev_size_mb=128
typeset -a disks
embedded_slog_min_ms=$(get_tunable EMBEDDED_SLOG_MIN_MS)
original_scrub_after_expand=$(get_tunable SCRUB_AFTER_EXPAND)
function cleanup
{
poolexists "$TESTPOOL" && log_must_busy zpool destroy "$TESTPOOL"
for i in {0..$devs}; do
log_must rm -f "$TEST_BASE_DIR/dev-$i"
done
log_must set_tunable32 EMBEDDED_SLOG_MIN_MS $embedded_slog_min_ms
log_must set_tunable32 SCRUB_AFTER_EXPAND $original_scrub_after_expand
}
log_onexit cleanup
log_must set_tunable32 EMBEDDED_SLOG_MIN_MS 99999
# Disk files which will be used by pool
for i in {0..$(($devs))}; do
device=$TEST_BASE_DIR/dev-$i
log_must truncate -s ${dev_size_mb}M $device
disks[${#disks[*]}+1]=$device
done
nparity=$((RANDOM%(3) + 1))
raid=raidz$nparity
pool=$TESTPOOL
opts="-o cachefile=none"
log_must set_tunable32 SCRUB_AFTER_EXPAND 0
log_must zpool create -f $opts $pool $raid ${disks[1..$(($nparity+1))]}
log_must zfs create -o recordsize=8k $pool/fs
log_must fill_fs /$pool/fs 1 128 100 1024 R
log_must zfs create -o recordsize=128k $pool/fs2
log_must fill_fs /$pool/fs2 1 128 100 1024 R
for disk in ${disks[$(($nparity+2))..$devs]}; do
log_must zpool attach $pool ${raid}-0 $disk
sleep 10
for (( i=1; i<=$nparity; i=i+1 )); do
log_must zpool offline $pool ${disks[$i]}
log_must dd if=/dev/zero of=${disks[$i]} \
bs=1024k count=$dev_size_mb conv=notrunc
done
sleep 3
for (( i=1; i<=$nparity; i=i+1 )); do
log_must zpool replace $pool ${disks[$i]}
done
log_must zpool wait -t replace $pool
log_must check_pool_status $pool "scan" "with 0 errors"
log_must zpool wait -t raidz_expand $pool
log_must zpool clear $pool
log_must zpool scrub -w $pool
log_must zpool status -v
log_must check_pool_status $pool "scan" "with 0 errors"
done
log_must zpool destroy "$pool"
log_pass "raidz expansion test succeeded."
@@ -0,0 +1,177 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2021 by vStack. All rights reserved.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# Check device replacement during raidz expansion using expansion pausing.
#
# STRATEGY:
# 1. Create block device files for the test raidz pool
# 2. For each parity value [1..3]
# - create raidz pool with minimum block device files required
# - create couple of datasets with different recordsize and fill it
# - set raidz expand maximum reflow bytes
# - attach new device to the pool
# - wait for reflow bytes to reach the maximum
# - offline and zero vdevs allowed by parity
# - wait some time and start offlined vdevs replacement
# - wait replacement completion and verify pool status
# - loop thru vdevs replacing with the max reflow bytes increasing
# - verify pool
# - set reflow bytes to max value to complete the expansion
typeset -r devs=10
typeset -r dev_size_mb=128
typeset -a disks
embedded_slog_min_ms=$(get_tunable EMBEDDED_SLOG_MIN_MS)
original_scrub_after_expand=$(get_tunable SCRUB_AFTER_EXPAND)
function cleanup
{
poolexists "$TESTPOOL" && zpool status -v "$TESTPOOL"
poolexists "$TESTPOOL" && log_must_busy zpool destroy "$TESTPOOL"
for i in {0..$devs}; do
log_must rm -f "$TEST_BASE_DIR/dev-$i"
done
log_must set_tunable32 EMBEDDED_SLOG_MIN_MS $embedded_slog_min_ms
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES 0
log_must set_tunable32 SCRUB_AFTER_EXPAND $original_scrub_after_expand
}
function wait_expand_paused
{
oldcopied='0'
newcopied='1'
while [[ $oldcopied != $newcopied ]]; do
oldcopied=$newcopied
sleep 1
newcopied=$(zpool status $TESTPOOL | \
grep 'copied out of' | \
awk '{print $1}')
done
}
log_onexit cleanup
function test_replace # <pool> <devices> <parity>
{
pool=${1}
devices=${2}
nparity=${3}
device_count=0
log_must echo "devices=$devices"
for dev in ${devices}; do
device_count=$((device_count+1))
done
index=$((RANDOM%(device_count-nparity)))
for (( j=1; j<=$nparity; j=j+1 )); do
log_must zpool offline $pool ${disks[$((index+j))]}
log_must dd if=/dev/zero of=${disks[$((index+j))]} \
bs=1024k count=$dev_size_mb conv=notrunc
done
for (( j=1; j<=$nparity; j=j+1 )); do
log_must zpool replace $pool ${disks[$((index+j))]}
done
log_must zpool wait -t replace $pool
log_must check_pool_status $pool "scan" "with 0 errors"
log_must zpool clear $pool
log_must zpool scrub -w $pool
log_must zpool status -v
log_must check_pool_status $pool "scan" "repaired 0B"
}
log_must set_tunable32 EMBEDDED_SLOG_MIN_MS 99999
# Disk files which will be used by pool
for i in {0..$(($devs))}; do
device=$TEST_BASE_DIR/dev-$i
log_must truncate -s ${dev_size_mb}M $device
disks[${#disks[*]}+1]=$device
done
nparity=$((RANDOM%(3) + 1))
raid=raidz$nparity
pool=$TESTPOOL
opts="-o cachefile=none"
devices=""
log_must set_tunable32 SCRUB_AFTER_EXPAND 0
log_must zpool create -f $opts $pool $raid ${disks[1..$(($nparity+1))]}
devices="${disks[1..$(($nparity+1))]}"
log_must zfs create -o recordsize=8k $pool/fs
log_must fill_fs /$pool/fs 1 128 100 1024 R
log_must zfs create -o recordsize=128k $pool/fs2
log_must fill_fs /$pool/fs2 1 128 100 1024 R
for disk in ${disks[$(($nparity+2))..$devs]}; do
# Set pause to some random value near halfway point
reflow_size=$(get_pool_prop allocated $pool)
pause=$((((RANDOM << 15) + RANDOM) % reflow_size / 2))
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES $pause
log_must zpool attach $pool ${raid}-0 $disk
devices="$devices $disk"
wait_expand_paused
for (( i=0; i<2; i++ )); do
test_replace $pool "$devices" $nparity
# Increase pause by about 25%
pause=$((pause + (((RANDOM << 15) + RANDOM) % \
reflow_size) / 4))
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES $pause
wait_expand_paused
done
# Set pause past largest possible value for this pool
pause=$((devs*dev_size_mb*1024*1024))
log_must set_tunable64 RAIDZ_EXPAND_MAX_REFLOW_BYTES $pause
log_must zpool wait -t raidz_expand $pool
done
log_must zpool destroy "$pool"
log_pass "raidz expansion test succeeded."
@@ -0,0 +1,78 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2023 by iXsystems, Inc.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# 'zpool attach poolname raidz ...' should fail if raidz_expansion
# feature is not enabled.
#
# STRATEGY:
# 1. Create raidz pool with raidz_expansion feature disabled
# 2. Attempt to attach a device to the raidz vdev
# 3. Verify that device attached failed
# 4. Destroy the raidz pool
typeset -r devs=4
typeset -r dev_size_mb=128
typeset -a disks
function cleanup
{
log_pos zpool status "$TESTPOOL"
poolexists "$TESTPOOL" && log_must_busy zpool destroy "$TESTPOOL"
for i in {0..$devs}; do
log_must rm -f "$TEST_BASE_DIR/dev-$i"
done
}
log_onexit cleanup
for i in {0..$devs}; do
device=$TEST_BASE_DIR/dev-$i
log_must truncate -s ${dev_size_mb}M "$device"
if [[ $i -ne $devs ]]; then
disks[${#disks[*]}+1]=$device
fi
done
# create a pool with raidz_expansion feature disabled
log_must zpool create -f -o cachefile=none -o feature@raidz_expansion=disabled \
"$TESTPOOL" raidz1 "${disks[@]}"
status=$(zpool list -H -o feature@raidz_expansion "$TESTPOOL")
if [[ "$status" != "disabled" ]]; then
log_fail "raidz_expansion feature was not disabled"
fi
# expecting attach to fail
log_mustnot_expect "raidz_expansion feature must be enabled" zpool attach -f \
"$TESTPOOL" raidz1-0 "$TEST_BASE_DIR/dev-$devs"
log_must zpool destroy "$TESTPOOL"
log_pass "raidz attach failed with feature disabled as expected"
@@ -0,0 +1,86 @@
#!/bin/ksh -p
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License (the "License").
# You may not use this file except in compliance with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright (c) 2023 by iXsystems, Inc.
#
. $STF_SUITE/include/libtest.shlib
#
# DESCRIPTION:
# Negative for FreeBSD Only
#
# Attempting to expand a RAIDZ should fail if the scratch area on the
# existing disks contains BTX Server binary (used to boot FreeBSD when
# using MBR partitions with ZFS).
#
# STRATEGY:
# 1. Create raidz pool
# 2. Add a BTX header to the reserved boot area
# 3. Attempt to attach a device to the raidz vdev
# 4. Verify that device attached failed
# 5. Destroy the raidz pool
typeset -r devs=4
typeset -r dev_size_mb=128
typeset -a disks
function cleanup
{
log_pos zpool status "$TESTPOOL"
poolexists "$TESTPOOL" && log_must_busy zpool destroy "$TESTPOOL"
for i in {0..$devs}; do
log_must rm -f "$TEST_BASE_DIR/dev-$i"
done
}
log_onexit cleanup
for i in {0..$devs}; do
device=$TEST_BASE_DIR/dev-$i
# simulate active BTX Server data by inserting a BTX header
printf "\xeb\x0e%s\x01\x02\x80" "BTX" | dd of="$device" \
bs=512 seek=1024 status=none
log_must truncate -s ${dev_size_mb}M "$device"
if [[ $i -ne $devs ]]; then
disks[${#disks[*]}+1]=$device
fi
done
log_must zpool create -f -o cachefile=none "$TESTPOOL" raidz1 "${disks[@]}"
if is_freebsd; then
# expecting attach to fail
log_mustnot_expect "the reserved boot area" zpool attach -f \
"$TESTPOOL" raidz1-0 "$TEST_BASE_DIR/dev-$devs"
log_must zpool destroy "$TESTPOOL"
log_pass "raidz attach failed with in-use reserved boot area"
else
# expecting attach to pass everywhere else
log_must zpool attach -f "$TESTPOOL" raidz1-0 "$TEST_BASE_DIR/dev-$devs"
log_must zpool destroy "$TESTPOOL"
log_pass "raidz attach passed with in-use reserved boot area"
fi