freebsd-dev/scripts/zconfig.sh

675 lines
20 KiB
Bash
Raw Normal View History

#!/bin/bash
#
# ZFS/ZPOOL configuration test script.
basedir="$(dirname $0)"
SCRIPT_COMMON=common.sh
if [ -f "${basedir}/${SCRIPT_COMMON}" ]; then
. "${basedir}/${SCRIPT_COMMON}"
else
echo "Missing helper script ${SCRIPT_COMMON}" && exit 1
fi
PROG=zconfig.sh
usage() {
cat << EOF
USAGE:
$0 [hvcts]
DESCRIPTION:
ZFS/ZPOOL configuration tests
OPTIONS:
-h Show this message
-v Verbose
-c Cleanup lo+file devices at start
-t <#> Run listed tests
-s <#> Skip listed tests
EOF
}
while getopts 'hvct:s:?' OPTION; do
case $OPTION in
h)
usage
exit 1
;;
v)
VERBOSE=1
;;
c)
CLEANUP=1
;;
t)
TESTS_RUN=($OPTARG)
;;
s)
TESTS_SKIP=($OPTARG)
;;
?)
usage
exit
;;
esac
done
if [ $(id -u) != 0 ]; then
die "Must run as root"
fi
# Initialize the test suite
init
# Perform pre-cleanup is requested
if [ ${CLEANUP} ]; then
${ZFS_SH} -u
Add zfault zpool configurations and tests Eleven new zpool configurations were added to allow testing of various failure cases. The first 5 zpool configurations leverage the 'faulty' md device type which allow us to simuluate IO errors at the block layer. The last 6 zpool configurations leverage the scsi_debug module provided by modern kernels. This device allows you to create virtual scsi devices which are backed by a ram disk. With this setup we can verify the full IO stack by injecting faults at the lowest layer. Both methods of fault injection are important to verifying the IO stack. The zfs code itself also provides a mechanism for error injection via the zinject command line tool. While we should also take advantage of this appraoch to validate the code it does not address any of the Linux integration issues which are the most concerning. For the moment we're trusting that the upstream Solaris guys are running zinject and would have caught internal zfs logic errors. Currently, there are 6 r/w test cases layered on top of the 'faulty' md devices. They include 3 writes tests for soft/transient errors, hard/permenant errors, and all writes error to the device. There are 3 matching read tests for soft/transient errors, hard/permenant errors, and fixable read error with a write. Although for this last case zfs doesn't do anything special. The seventh test case verifies zfs detects and corrects checksum errors. In this case one of the drives is extensively damaged and by dd'ing over large sections of it. We then ensure zfs logs the issue and correctly rebuilds the damage. The next test cases use the scsi_debug configuration to injects error at the bottom of the scsi stack. This ensures we find any flaws in the scsi midlayer or our usage of it. Plus it stresses the device specific retry, timeout, and error handling outside of zfs's control. The eighth test case is to verify that the system correctly handles an intermittent device timeout. Here the scsi_debug device drops 1 in N requests resulting in a retry either at the block level. The ZFS code does specify the FAILFAST option but it turns out that for this case the Linux IO stack with still retry the command. The FAILFAST logic located in scsi_noretry_cmd() does no seem to apply to the simply timeout case. It appears to be more targeted to specific device or transport errors from the lower layers. The ninth test case handles a persistent failure in which the device is removed from the system by Linux. The test verifies that the failure is detected, the device is made unavailable, and then can be successfully re-add when brought back online. Additionally, it ensures that errors and events are logged to the correct places and the no data corruption has occured due to the failure.
2010-09-28 23:32:12 +00:00
cleanup_md_devices
cleanup_loop_devices
rm -f /tmp/zpool.cache.*
fi
# Check if we need to skip the tests that require scsi_debug and lsscsi.
SCSI_DEBUG=0
${INFOMOD} scsi_debug &>/dev/null && SCSI_DEBUG=1
HAVE_LSSCSI=0
test -f ${LSSCSI} && HAVE_LSSCSI=1
if [ ${SCSI_DEBUG} -eq 0 ] || [ ${HAVE_LSSCSI} -eq 0 ]; then
echo "Skipping test 10 which requires the scsi_debug " \
"module and the ${LSSCSI} utility"
fi
# Validate persistent zpool.cache configuration.
test_1() {
local POOL_NAME=test1
local TMP_FILE1=`mktemp`
local TMP_FILE2=`mktemp`
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
# Create a pool save its status for comparison.
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 || fail 2
${ZPOOL} status ${POOL_NAME} >${TMP_FILE1} || fail 3
# Unload/load the module stack and verify the pool persists.
${ZFS_SH} -u || fail 4
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 5
${ZPOOL} import -c ${TMP_CACHE} ${POOL_NAME} || fail 5
${ZPOOL} status ${POOL_NAME} >${TMP_FILE2} || fail 6
cmp ${TMP_FILE1} ${TMP_FILE2} || fail 7
# Cleanup the test pool and temporary files
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 8
rm -f ${TMP_FILE1} ${TMP_FILE2} ${TMP_CACHE} || fail 9
${ZFS_SH} -u || fail 10
pass
}
run_test 1 "persistent zpool.cache"
# Validate ZFS disk scanning and import w/out zpool.cache configuration.
test_2() {
local POOL_NAME=test2
local TMP_FILE1=`mktemp`
local TMP_FILE2=`mktemp`
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
# Create a pool save its status for comparison.
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 || fail 2
${ZPOOL} status ${POOL_NAME} >${TMP_FILE1} || fail 3
# Unload the module stack, remove the cache file, load the module
# stack and attempt to probe the disks to import the pool. As
# a cross check verify the old pool state against the imported.
${ZFS_SH} -u || fail 4
rm -f ${TMP_CACHE} || fail 5
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 6
${ZPOOL} import -d /dev ${POOL_NAME} || fail 8
${ZPOOL} status ${POOL_NAME} >${TMP_FILE2} || fail 9
cmp ${TMP_FILE1} ${TMP_FILE2} || fail 10
# Cleanup the test pool and temporary files
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 11
rm -f ${TMP_FILE1} ${TMP_FILE2} || fail 12
${ZFS_SH} -u || fail 13
pass
}
run_test 2 "scan disks for pools to import"
zconfig_zvol_device_stat() {
local EXPECT=$1
local POOL_NAME=/dev/zvol/$2
local ZVOL_NAME=/dev/zvol/$3
local SNAP_NAME=/dev/zvol/$4
local CLONE_NAME=/dev/zvol/$5
local COUNT=0
# Briefly delay for udev
udev_trigger
# Pool exists
stat ${POOL_NAME} &>/dev/null && let COUNT=$COUNT+1
# Volume and partitions
stat ${ZVOL_NAME} &>/dev/null && let COUNT=$COUNT+1
stat ${ZVOL_NAME}-part1 &>/dev/null && let COUNT=$COUNT+1
stat ${ZVOL_NAME}-part2 &>/dev/null && let COUNT=$COUNT+1
# Snapshot with partitions
stat ${SNAP_NAME} &>/dev/null && let COUNT=$COUNT+1
stat ${SNAP_NAME}-part1 &>/dev/null && let COUNT=$COUNT+1
stat ${SNAP_NAME}-part2 &>/dev/null && let COUNT=$COUNT+1
# Clone with partitions
stat ${CLONE_NAME} &>/dev/null && let COUNT=$COUNT+1
stat ${CLONE_NAME}-part1 &>/dev/null && let COUNT=$COUNT+1
stat ${CLONE_NAME}-part2 &>/dev/null && let COUNT=$COUNT+1
if [ $EXPECT -ne $COUNT ]; then
return 1
fi
return 0
}
# zpool import/export device check
# (1 volume, 2 partitions, 1 snapshot, 1 clone)
test_3() {
local POOL_NAME=tank
local ZVOL_NAME=volume
local SNAP_NAME=snap
local CLONE_NAME=clone
local FULL_ZVOL_NAME=${POOL_NAME}/${ZVOL_NAME}
local FULL_SNAP_NAME=${POOL_NAME}/${ZVOL_NAME}@${SNAP_NAME}
local FULL_CLONE_NAME=${POOL_NAME}/${CLONE_NAME}
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
# Create a pool, volume, partition, snapshot, and clone.
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 || fail 2
${ZFS} create -V 100M ${FULL_ZVOL_NAME} || fail 3
Add snapdev=[hidden|visible] dataset property The new snapdev dataset property may be set to control the visibility of zvol snapshot devices. By default this value is set to 'hidden' which will prevent zvol snapshots from appearing under /dev/zvol/ and /dev/<dataset>/. When set to 'visible' all zvol snapshots for the dataset will be visible. This functionality was largely added because when automatic snapshoting is enabled large numbers of read-only zvol snapshots will be created. When creating these devices the kernel will attempt to read their partition tables, and blkid will attempt to identify any filesystems on those partitions. This leads to a variety of issues: 1) The zvol partition tables will be read in the context of the `modprobe zfs` for automatically imported pools. This is undesirable and should be done asynchronously, but for now reducing the number of visible devices helps. 2) Udev expects to be able to complete its work for a new block devices fairly quickly. When many zvol devices are added at the same time this is no longer be true. It can lead to udev timeouts and missing /dev/zvol links. 3) Simply having lots of devices in /dev/ can be aukward from a management standpoint. Hidding the devices your unlikely to ever use helps with this. Any snapshot device which is needed can be made visible by changing the snapdev property. NOTE: This patch changes the default behavior for zvols which was effectively 'snapdev=visible'. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #1235 Closes #945 Issue #956 Issue #756
2013-02-13 23:11:59 +00:00
${ZFS} set snapdev=visible ${FULL_ZVOL_NAME} || fail 3
label /dev/zvol/${FULL_ZVOL_NAME} msdos || fail 4
partition /dev/zvol/${FULL_ZVOL_NAME} primary 1% 50% || fail 4
partition /dev/zvol/${FULL_ZVOL_NAME} primary 51% -1 || fail 4
${ZFS} snapshot ${FULL_SNAP_NAME} || fail 5
${ZFS} clone ${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 6
# Verify the devices were created
zconfig_zvol_device_stat 10 ${POOL_NAME} ${FULL_ZVOL_NAME} \
${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 7
# Export the pool
${ZPOOL} export ${POOL_NAME} || fail 8
# verify the devices were removed
zconfig_zvol_device_stat 0 ${POOL_NAME} ${FULL_ZVOL_NAME} \
${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 9
# Import the pool, wait 1 second for udev
${ZPOOL} import ${POOL_NAME} || fail 10
# Verify the devices were created
zconfig_zvol_device_stat 10 ${POOL_NAME} ${FULL_ZVOL_NAME} \
${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 11
# Destroy the pool and consequently the devices
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 12
# verify the devices were removed
zconfig_zvol_device_stat 0 ${POOL_NAME} ${FULL_ZVOL_NAME} \
${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 13
${ZFS_SH} -u || fail 14
rm -f ${TMP_CACHE} || fail 15
pass
}
run_test 3 "zpool import/export device"
# zpool insmod/rmmod device check (1 volume, 1 snapshot, 1 clone)
test_4() {
POOL_NAME=tank
ZVOL_NAME=volume
SNAP_NAME=snap
CLONE_NAME=clone
FULL_ZVOL_NAME=${POOL_NAME}/${ZVOL_NAME}
FULL_SNAP_NAME=${POOL_NAME}/${ZVOL_NAME}@${SNAP_NAME}
FULL_CLONE_NAME=${POOL_NAME}/${CLONE_NAME}
TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
# Create a pool, volume, snapshot, and clone
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 || fail 2
${ZFS} create -V 100M ${FULL_ZVOL_NAME} || fail 3
Add snapdev=[hidden|visible] dataset property The new snapdev dataset property may be set to control the visibility of zvol snapshot devices. By default this value is set to 'hidden' which will prevent zvol snapshots from appearing under /dev/zvol/ and /dev/<dataset>/. When set to 'visible' all zvol snapshots for the dataset will be visible. This functionality was largely added because when automatic snapshoting is enabled large numbers of read-only zvol snapshots will be created. When creating these devices the kernel will attempt to read their partition tables, and blkid will attempt to identify any filesystems on those partitions. This leads to a variety of issues: 1) The zvol partition tables will be read in the context of the `modprobe zfs` for automatically imported pools. This is undesirable and should be done asynchronously, but for now reducing the number of visible devices helps. 2) Udev expects to be able to complete its work for a new block devices fairly quickly. When many zvol devices are added at the same time this is no longer be true. It can lead to udev timeouts and missing /dev/zvol links. 3) Simply having lots of devices in /dev/ can be aukward from a management standpoint. Hidding the devices your unlikely to ever use helps with this. Any snapshot device which is needed can be made visible by changing the snapdev property. NOTE: This patch changes the default behavior for zvols which was effectively 'snapdev=visible'. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #1235 Closes #945 Issue #956 Issue #756
2013-02-13 23:11:59 +00:00
${ZFS} set snapdev=visible ${FULL_ZVOL_NAME} || fail 3
label /dev/zvol/${FULL_ZVOL_NAME} msdos || fail 4
partition /dev/zvol/${FULL_ZVOL_NAME} primary 1% 50% || fail 4
partition /dev/zvol/${FULL_ZVOL_NAME} primary 51% -1 || fail 4
${ZFS} snapshot ${FULL_SNAP_NAME} || fail 5
${ZFS} clone ${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 6
# Verify the devices were created
zconfig_zvol_device_stat 10 ${POOL_NAME} ${FULL_ZVOL_NAME} \
${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 7
# Unload the modules
${ZFS_SH} -u || fail 8
# Verify the devices were removed
zconfig_zvol_device_stat 0 ${POOL_NAME} ${FULL_ZVOL_NAME} \
${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 9
# Load the modules, list the pools to ensure they are opened
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 10
${ZPOOL} import -c ${TMP_CACHE} ${POOL_NAME} || fail 10
${ZPOOL} list &>/dev/null
# Verify the devices were created
zconfig_zvol_device_stat 10 ${POOL_NAME} ${FULL_ZVOL_NAME} \
${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 11
# Destroy the pool and consequently the devices
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 12
# Verify the devices were removed
zconfig_zvol_device_stat 0 ${POOL_NAME} ${FULL_ZVOL_NAME} \
${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 13
${ZFS_SH} -u || fail 14
rm -f ${TMP_CACHE} || fail 15
pass
}
run_test 4 "zpool insmod/rmmod device"
# ZVOL volume sanity check
test_5() {
local POOL_NAME=tank
local ZVOL_NAME=fish
local FULL_NAME=${POOL_NAME}/${ZVOL_NAME}
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
# Create a pool and volume.
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raid0 || fail 2
${ZFS} create -V 800M ${FULL_NAME} || fail 3
label /dev/zvol/${FULL_NAME} msdos || fail 4
partition /dev/zvol/${FULL_NAME} primary 1 -1 || fail 4
format /dev/zvol/${FULL_NAME}-part1 ext2 || fail 5
# Mount the ext2 filesystem and copy some data to it.
mkdir -p /tmp/${ZVOL_NAME}-part1 || fail 6
mount /dev/zvol/${FULL_NAME}-part1 /tmp/${ZVOL_NAME}-part1 || fail 7
cp -RL ${SRC_DIR} /tmp/${ZVOL_NAME}-part1 || fail 8
sync
# Verify the copied files match the original files.
diff -ur ${SRC_DIR} /tmp/${ZVOL_NAME}-part1/${SRC_DIR##*/} \
&>/dev/null || fail 9
# Remove the files, umount, destroy the volume and pool.
rm -Rf /tmp/${ZVOL_NAME}-part1/${SRC_DIR##*/} || fail 10
umount /tmp/${ZVOL_NAME}-part1 || fail 11
rmdir /tmp/${ZVOL_NAME}-part1 || fail 12
${ZFS} destroy ${FULL_NAME} || fail 13
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 14
${ZFS_SH} -u || fail 15
rm -f ${TMP_CACHE} || fail 16
pass
}
run_test 5 "zvol+ext2 volume"
# ZVOL snapshot sanity check
test_6() {
local POOL_NAME=tank
local ZVOL_NAME=fish
local SNAP_NAME=pristine
local FULL_ZVOL_NAME=${POOL_NAME}/${ZVOL_NAME}
local FULL_SNAP_NAME=${POOL_NAME}/${ZVOL_NAME}@${SNAP_NAME}
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
# Create a pool and volume.
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raid0 || fail 2
${ZFS} create -V 800M ${FULL_ZVOL_NAME} || fail 3
Add snapdev=[hidden|visible] dataset property The new snapdev dataset property may be set to control the visibility of zvol snapshot devices. By default this value is set to 'hidden' which will prevent zvol snapshots from appearing under /dev/zvol/ and /dev/<dataset>/. When set to 'visible' all zvol snapshots for the dataset will be visible. This functionality was largely added because when automatic snapshoting is enabled large numbers of read-only zvol snapshots will be created. When creating these devices the kernel will attempt to read their partition tables, and blkid will attempt to identify any filesystems on those partitions. This leads to a variety of issues: 1) The zvol partition tables will be read in the context of the `modprobe zfs` for automatically imported pools. This is undesirable and should be done asynchronously, but for now reducing the number of visible devices helps. 2) Udev expects to be able to complete its work for a new block devices fairly quickly. When many zvol devices are added at the same time this is no longer be true. It can lead to udev timeouts and missing /dev/zvol links. 3) Simply having lots of devices in /dev/ can be aukward from a management standpoint. Hidding the devices your unlikely to ever use helps with this. Any snapshot device which is needed can be made visible by changing the snapdev property. NOTE: This patch changes the default behavior for zvols which was effectively 'snapdev=visible'. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #1235 Closes #945 Issue #956 Issue #756
2013-02-13 23:11:59 +00:00
${ZFS} set snapdev=visible ${FULL_ZVOL_NAME} || fail 3
label /dev/zvol/${FULL_ZVOL_NAME} msdos || fail 4
partition /dev/zvol/${FULL_ZVOL_NAME} primary 1 -1 || fail 4
format /dev/zvol/${FULL_ZVOL_NAME}-part1 ext2 || fail 5
# Mount the ext2 filesystem and copy some data to it.
mkdir -p /tmp/${ZVOL_NAME}-part1 || fail 6
mount /dev/zvol/${FULL_ZVOL_NAME}-part1 /tmp/${ZVOL_NAME}-part1 \
|| fail 7
# Snapshot the pristine ext2 filesystem and mount it read-only.
${ZFS} snapshot ${FULL_SNAP_NAME} || fail 8
wait_udev /dev/zvol/${FULL_SNAP_NAME}-part1 30 || fail 8
mkdir -p /tmp/${SNAP_NAME}-part1 || fail 9
mount /dev/zvol/${FULL_SNAP_NAME}-part1 /tmp/${SNAP_NAME}-part1 \
&>/dev/null || fail 10
# Copy to original volume
cp -RL ${SRC_DIR} /tmp/${ZVOL_NAME}-part1 || fail 11
sync
# Verify the copied files match the original files,
# and the copied files do NOT appear in the snapshot.
diff -ur ${SRC_DIR} /tmp/${ZVOL_NAME}-part1/${SRC_DIR##*/} \
&>/dev/null || fail 12
diff -ur ${SRC_DIR} /tmp/${SNAP_NAME}-part1/${SRC_DIR##*/} \
&>/dev/null && fail 13
# umount, destroy the snapshot, volume, and pool.
umount /tmp/${SNAP_NAME}-part1 || fail 14
rmdir /tmp/${SNAP_NAME}-part1 || fail 15
${ZFS} destroy ${FULL_SNAP_NAME} || fail 16
umount /tmp/${ZVOL_NAME}-part1 || fail 17
rmdir /tmp/${ZVOL_NAME}-part1 || fail 18
${ZFS} destroy ${FULL_ZVOL_NAME} || fail 19
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 20
${ZFS_SH} -u || fail 21
rm -f ${TMP_CACHE} || fail 22
pass
}
run_test 6 "zvol+ext2 snapshot"
# ZVOL clone sanity check
test_7() {
local POOL_NAME=tank
local ZVOL_NAME=fish
local SNAP_NAME=pristine
local CLONE_NAME=clone
local FULL_ZVOL_NAME=${POOL_NAME}/${ZVOL_NAME}
local FULL_SNAP_NAME=${POOL_NAME}/${ZVOL_NAME}@${SNAP_NAME}
local FULL_CLONE_NAME=${POOL_NAME}/${CLONE_NAME}
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
# Create a pool and volume.
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 || fail 2
${ZFS} create -V 300M ${FULL_ZVOL_NAME} || fail 3
Add snapdev=[hidden|visible] dataset property The new snapdev dataset property may be set to control the visibility of zvol snapshot devices. By default this value is set to 'hidden' which will prevent zvol snapshots from appearing under /dev/zvol/ and /dev/<dataset>/. When set to 'visible' all zvol snapshots for the dataset will be visible. This functionality was largely added because when automatic snapshoting is enabled large numbers of read-only zvol snapshots will be created. When creating these devices the kernel will attempt to read their partition tables, and blkid will attempt to identify any filesystems on those partitions. This leads to a variety of issues: 1) The zvol partition tables will be read in the context of the `modprobe zfs` for automatically imported pools. This is undesirable and should be done asynchronously, but for now reducing the number of visible devices helps. 2) Udev expects to be able to complete its work for a new block devices fairly quickly. When many zvol devices are added at the same time this is no longer be true. It can lead to udev timeouts and missing /dev/zvol links. 3) Simply having lots of devices in /dev/ can be aukward from a management standpoint. Hidding the devices your unlikely to ever use helps with this. Any snapshot device which is needed can be made visible by changing the snapdev property. NOTE: This patch changes the default behavior for zvols which was effectively 'snapdev=visible'. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #1235 Closes #945 Issue #956 Issue #756
2013-02-13 23:11:59 +00:00
${ZFS} set snapdev=visible ${FULL_ZVOL_NAME} || fail 3
label /dev/zvol/${FULL_ZVOL_NAME} msdos || fail 4
partition /dev/zvol/${FULL_ZVOL_NAME} primary 1 -1 || fail 4
format /dev/zvol/${FULL_ZVOL_NAME}-part1 ext2 || fail 5
# Snapshot the pristine ext2 filesystem.
${ZFS} snapshot ${FULL_SNAP_NAME} || fail 6
wait_udev /dev/zvol/${FULL_SNAP_NAME}-part1 30 || fail 7
# Mount the ext2 filesystem so some data can be copied to it.
mkdir -p /tmp/${ZVOL_NAME}-part1 || fail 7
mount /dev/zvol/${FULL_ZVOL_NAME}-part1 \
/tmp/${ZVOL_NAME}-part1 || fail 8
# Mount the pristine ext2 snapshot.
mkdir -p /tmp/${SNAP_NAME}-part1 || fail 9
mount /dev/zvol/${FULL_SNAP_NAME}-part1 \
/tmp/${SNAP_NAME}-part1 &>/dev/null || fail 10
# Copy to original volume.
cp -RL ${SRC_DIR} /tmp/${ZVOL_NAME}-part1 || fail 11
sync
# Verify the copied files match the original files,
# and the copied files do NOT appear in the snapshot.
diff -ur ${SRC_DIR} /tmp/${ZVOL_NAME}-part1/${SRC_DIR##*/} \
&>/dev/null || fail 12
diff -ur ${SRC_DIR} /tmp/${SNAP_NAME}-part1/${SRC_DIR##*/} \
&>/dev/null && fail 13
# Clone from the original pristine snapshot
${ZFS} clone ${FULL_SNAP_NAME} ${FULL_CLONE_NAME} || fail 14
wait_udev /dev/zvol/${FULL_CLONE_NAME}-part1 30 || fail 14
mkdir -p /tmp/${CLONE_NAME}-part1 || fail 15
mount /dev/zvol/${FULL_CLONE_NAME}-part1 \
/tmp/${CLONE_NAME}-part1 || fail 16
# Verify the clone matches the pristine snapshot,
# and the files copied to the original volume are NOT there.
diff -ur /tmp/${SNAP_NAME}-part1 /tmp/${CLONE_NAME}-part1 \
&>/dev/null || fail 17
diff -ur /tmp/${ZVOL_NAME}-part1 /tmp/${CLONE_NAME}-part1 \
&>/dev/null && fail 18
# Copy to cloned volume.
cp -RL ${SRC_DIR} /tmp/${CLONE_NAME}-part1 || fail 19
sync
# Verify the clone matches the modified original volume.
diff -ur /tmp/${ZVOL_NAME}-part1 /tmp/${CLONE_NAME}-part1 \
&>/dev/null || fail 20
# umount, destroy the snapshot, volume, and pool.
umount /tmp/${CLONE_NAME}-part1 || fail 21
rmdir /tmp/${CLONE_NAME}-part1 || fail 22
${ZFS} destroy ${FULL_CLONE_NAME} || fail 23
umount /tmp/${SNAP_NAME}-part1 || fail 24
rmdir /tmp/${SNAP_NAME}-part1 || fail 25
${ZFS} destroy ${FULL_SNAP_NAME} || fail 26
umount /tmp/${ZVOL_NAME}-part1 || fail 27
rmdir /tmp/${ZVOL_NAME}-part1 || fail 28
${ZFS} destroy ${FULL_ZVOL_NAME} || fail 29
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 30
${ZFS_SH} -u || fail 31
rm -f ${TMP_CACHE} || fail 32
pass
}
run_test 7 "zvol+ext2 clone"
# Send/Receive sanity check
test_8() {
local POOL_NAME1=tank1
local POOL_NAME2=tank2
local ZVOL_NAME=fish
local SNAP_NAME=snap
local FULL_ZVOL_NAME1=${POOL_NAME1}/${ZVOL_NAME}
local FULL_ZVOL_NAME2=${POOL_NAME2}/${ZVOL_NAME}
local FULL_SNAP_NAME1=${POOL_NAME1}/${ZVOL_NAME}@${SNAP_NAME}
local FULL_SNAP_NAME2=${POOL_NAME2}/${ZVOL_NAME}@${SNAP_NAME}
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
# Create two pools and a volume
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME1} -c lo-raidz2 || fail 2
${ZPOOL_CREATE_SH} -p ${POOL_NAME2} -c lo-raidz2 || fail 2
${ZFS} create -V 300M ${FULL_ZVOL_NAME1} || fail 3
Add snapdev=[hidden|visible] dataset property The new snapdev dataset property may be set to control the visibility of zvol snapshot devices. By default this value is set to 'hidden' which will prevent zvol snapshots from appearing under /dev/zvol/ and /dev/<dataset>/. When set to 'visible' all zvol snapshots for the dataset will be visible. This functionality was largely added because when automatic snapshoting is enabled large numbers of read-only zvol snapshots will be created. When creating these devices the kernel will attempt to read their partition tables, and blkid will attempt to identify any filesystems on those partitions. This leads to a variety of issues: 1) The zvol partition tables will be read in the context of the `modprobe zfs` for automatically imported pools. This is undesirable and should be done asynchronously, but for now reducing the number of visible devices helps. 2) Udev expects to be able to complete its work for a new block devices fairly quickly. When many zvol devices are added at the same time this is no longer be true. It can lead to udev timeouts and missing /dev/zvol links. 3) Simply having lots of devices in /dev/ can be aukward from a management standpoint. Hidding the devices your unlikely to ever use helps with this. Any snapshot device which is needed can be made visible by changing the snapdev property. NOTE: This patch changes the default behavior for zvols which was effectively 'snapdev=visible'. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #1235 Closes #945 Issue #956 Issue #756
2013-02-13 23:11:59 +00:00
${ZFS} set snapdev=visible ${FULL_ZVOL_NAME1} || fail 3
label /dev/zvol/${FULL_ZVOL_NAME1} msdos || fail 4
partition /dev/zvol/${FULL_ZVOL_NAME1} primary 1 -1 || fail 4
format /dev/zvol/${FULL_ZVOL_NAME1}-part1 ext2 || fail 5
# Mount the ext2 filesystem and copy some data to it.
mkdir -p /tmp/${FULL_ZVOL_NAME1}-part1 || fail 6
mount /dev/zvol/${FULL_ZVOL_NAME1}-part1 \
/tmp/${FULL_ZVOL_NAME1}-part1 || fail 7
cp -RL ${SRC_DIR} /tmp/${FULL_ZVOL_NAME1}-part1 || fail 8
# Unmount, snapshot, mount the ext2 filesystem so it may be sent.
# We only unmount to ensure the ext2 filesystem is clean.
umount /tmp/${FULL_ZVOL_NAME1}-part1 || fail 9
${ZFS} snapshot ${FULL_SNAP_NAME1} || fail 10
wait_udev /dev/zvol/${FULL_SNAP_NAME1} 30 || fail 10
mount /dev/zvol/${FULL_ZVOL_NAME1}-part1 \
/tmp/${FULL_ZVOL_NAME1}-part1 || 11
# Send/receive the snapshot from POOL_NAME1 to POOL_NAME2
(${ZFS} send ${FULL_SNAP_NAME1} | \
${ZFS} receive ${FULL_ZVOL_NAME2}) || fail 12
wait_udev /dev/zvol/${FULL_ZVOL_NAME2}-part1 30 || fail 12
# Mount the sent ext2 filesystem.
mkdir -p /tmp/${FULL_ZVOL_NAME2}-part1 || fail 13
mount /dev/zvol/${FULL_ZVOL_NAME2}-part1 \
/tmp/${FULL_ZVOL_NAME2}-part1 || fail 14
# Verify the contents of the volumes match
diff -ur /tmp/${FULL_ZVOL_NAME1}-part1 /tmp/${FULL_ZVOL_NAME2}-part1 \
&>/dev/null || fail 15
# Umount, destroy the volume and pool.
umount /tmp/${FULL_ZVOL_NAME1}-part1 || fail 16
umount /tmp/${FULL_ZVOL_NAME2}-part1 || fail 17
rmdir /tmp/${FULL_ZVOL_NAME1}-part1 || fail 18
rmdir /tmp/${FULL_ZVOL_NAME2}-part1 || fail 19
rmdir /tmp/${POOL_NAME1} || fail 20
rmdir /tmp/${POOL_NAME2} || fail 21
${ZFS} destroy ${FULL_SNAP_NAME1} || fail 22
${ZFS} destroy ${FULL_SNAP_NAME2} || fail 23
${ZFS} destroy ${FULL_ZVOL_NAME1} || fail 24
${ZFS} destroy ${FULL_ZVOL_NAME2} || fail 25
${ZPOOL_CREATE_SH} -p ${POOL_NAME1} -c lo-raidz2 -d || fail 26
${ZPOOL_CREATE_SH} -p ${POOL_NAME2} -c lo-raidz2 -d || fail 27
${ZFS_SH} -u || fail 28
rm -f ${TMP_CACHE} || fail 29
pass
}
run_test 8 "zfs send/receive"
# zpool event sanity check
test_9() {
local POOL_NAME=tank
local ZVOL_NAME=fish
local FULL_NAME=${POOL_NAME}/${ZVOL_NAME}
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
local TMP_EVENTS=`mktemp -p /tmp zpool.events.XXXXXXXX`
# Create a pool and volume.
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 || fail 2
${ZFS} create -V 300M ${FULL_NAME} || fail 3
udev_trigger
# Dump the events, there should be at least 5 lines.
${ZPOOL} events >${TMP_EVENTS} || fail 4
EVENTS=`wc -l ${TMP_EVENTS} | cut -f1 -d' '`
[ $EVENTS -lt 5 ] && fail 5
# Clear the events and ensure there are none.
${ZPOOL} events -c >/dev/null || fail 6
${ZPOOL} events >${TMP_EVENTS} || fail 7
EVENTS=`wc -l ${TMP_EVENTS} | cut -f1 -d' '`
[ $EVENTS -gt 1 ] && fail 8
${ZFS} destroy ${FULL_NAME} || fail 9
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 10
${ZFS_SH} -u || fail 11
rm -f ${TMP_CACHE} || fail 12
rm -f ${TMP_EVENTS} || fail 13
pass
}
run_test 9 "zpool events"
zconfig_add_vdev() {
local POOL_NAME=$1
local TYPE=$2
local DEVICE=$3
local TMP_FILE1=`mktemp`
local TMP_FILE2=`mktemp`
local TMP_FILE3=`mktemp`
BASE_DEVICE=`basename ${DEVICE}`
${ZPOOL} status ${POOL_NAME} >${TMP_FILE1}
${ZPOOL} add -f ${POOL_NAME} ${TYPE} ${DEVICE} 2>/dev/null || return 1
${ZPOOL} status ${POOL_NAME} >${TMP_FILE2}
diff ${TMP_FILE1} ${TMP_FILE2} > ${TMP_FILE3}
[ `wc -l ${TMP_FILE3}|${AWK} '{print $1}'` -eq 3 ] || return 1
PARENT_VDEV=`tail -2 ${TMP_FILE3} | head -1 | ${AWK} '{print $NF}'`
case $TYPE in
cache)
[ "${PARENT_VDEV}" = "${TYPE}" ] || return 1
;;
log)
[ "${PARENT_VDEV}" = "logs" ] || return 1
;;
esac
if ! tail -1 ${TMP_FILE3} |
egrep -q "^>[[:space:]]+${BASE_DEVICE}[[:space:]]+ONLINE" ; then
return 1
fi
rm -f ${TMP_FILE1} ${TMP_FILE2} ${TMP_FILE3}
return 0
}
# zpool add and remove sanity check
test_10() {
local POOL_NAME=tank
local TMP_CACHE=`mktemp -p /tmp zpool.cache.XXXXXXXX`
local TMP_FILE1=`mktemp`
local TMP_FILE2=`mktemp`
if [ ${SCSI_DEBUG} -eq 0 ] || [ ${HAVE_LSSCSI} -eq 0 ] ; then
skip
return
fi
test `${LSMOD} | grep -c scsi_debug` -gt 0 && \
(${RMMOD} scsi_debug || exit 1)
/sbin/modprobe scsi_debug dev_size_mb=128 ||
die "Error $? creating scsi_debug device"
udev_trigger
SDDEVICE=`${LSSCSI}|${AWK} '/scsi_debug/ { print $6; exit }'`
BASE_SDDEVICE=`basename $SDDEVICE`
# Create a pool
${ZFS_SH} zfs="spa_config_path=${TMP_CACHE}" || fail 1
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 || fail 2
${ZPOOL} status ${POOL_NAME} >${TMP_FILE1} || fail 3
# Add and remove a cache vdev by full path
zconfig_add_vdev ${POOL_NAME} cache ${SDDEVICE} || fail 4
${ZPOOL} remove ${POOL_NAME} ${SDDEVICE} || fail 5
${ZPOOL} status ${POOL_NAME} >${TMP_FILE2} || fail 6
cmp ${TMP_FILE1} ${TMP_FILE2} || fail 7
sleep 1
# Add and remove a cache vdev by shorthand path
zconfig_add_vdev ${POOL_NAME} cache ${BASE_SDDEVICE} || fail 8
${ZPOOL} remove ${POOL_NAME} ${BASE_SDDEVICE} || fail 9
${ZPOOL} status ${POOL_NAME} >${TMP_FILE2} || fail 10
cmp ${TMP_FILE1} ${TMP_FILE2} || fail 11
sleep 1
# Add and remove a log vdev
zconfig_add_vdev ${POOL_NAME} log ${BASE_SDDEVICE} || fail 12
${ZPOOL} remove ${POOL_NAME} ${BASE_SDDEVICE} || fail 13
${ZPOOL} status ${POOL_NAME} >${TMP_FILE2} || fail 14
cmp ${TMP_FILE1} ${TMP_FILE2} || fail 15
${ZPOOL_CREATE_SH} -p ${POOL_NAME} -c lo-raidz2 -d || fail 16
${ZFS_SH} -u || fail 17
${RMMOD} scsi_debug || fail 18
rm -f ${TMP_FILE1} ${TMP_FILE2} ${TMP_CACHE} || fail 19
pass
}
run_test 10 "zpool add/remove vdev"
exit 0