Re: [PATCH 2/3] xfstests: btrfs: test device replace, with EIO on the src dev

[Dave Chinner <david@fromorbit.com>]

On Thu, Apr 30, 2015 at 06:08:46PM +0800, Anand Jain wrote:
> This test case will test to confirm the replace works when
> the replacing source device has EIO errors.
> 
> EIO condition is achieved using the DM device.
> 
> Signed-off-by: Anand Jain <anand.jain@oracle.com>
> ---
>  tests/btrfs/087     | 97 +++++++++++++++++++++++++++++++++++++++++++++++++++++
>  tests/btrfs/087.out |  2 ++
>  tests/btrfs/group   |  1 +
>  3 files changed, 100 insertions(+)
>  create mode 100755 tests/btrfs/087
>  create mode 100644 tests/btrfs/087.out
> 
> diff --git a/tests/btrfs/087 b/tests/btrfs/087
> new file mode 100755
> index 0000000..c2eb16a
> --- /dev/null
> +++ b/tests/btrfs/087
> @@ -0,0 +1,97 @@
> +#! /bin/bash
> +# FS QA Test No. btrfs/087
> +#
> +# test device replace works when the source device has EIO
> +#
> +# Copyright (c) 2015 Oracle.  All Rights Reserved.
> +#
> +# This program is free software; you can redistribute it and/or
> +# modify it under the terms of the GNU General Public License as
> +# published by the Free Software Foundation.
> +#
> +# This program is distributed in the hope that it would be useful,
> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> +# GNU General Public License for more details.
> +#
> +# You should have received a copy of the GNU General Public License
> +# along with this program; if not, write the Free Software Foundation,
> +# Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
> +#
> +
> +seq=`basename $0`
> +seqres=$RESULT_DIR/$seq
> +echo "QA output created by $seq"
> +
> +here=`pwd`
> +tmp=/tmp/$$
> +status=1	# failure is the default!
> +trap "_cleanup; exit \$status" 0 1 2 3 15
> +rm -f $seqres.full
> +
> +_cleanup()
> +{
> +	# remove dm erro-dev from the scratch pool
> +	SCRATCH_DEV_POOL=`echo $SCRATCH_DEV_POOL | sed -e "s# *$DM_ERROR_DEV *##"`
> +
> +	_cleanup_dm_error_dev
> +	SCRATCH_DEV_POOL="$SCRATCH_DEV_POOL $DEV_REPLACE_SRC"
> +	SCRATCH_DEV_POOL="$SCRATCH_DEV_POOL $DEV_REPLACE_TGT"
> +	for i in $SCRATCH_DEV_POOL; do
> +		$WIPEFS_PROG -a $i > /dev/null 2>&1
> +	done

If you need to ensure the devices are clean, then that is an
intialisation job, not a cleanup job. Otherwise the automatic
scratch fs check will fail....

> +	rm -f $tmp.*
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +. ./common/dmerror
> +
> +_need_to_be_root
> +_supported_fs btrfs
> +_supported_os Linux
> +_require_scratch
> +_require_scratch_dev_pool 4
> +_require_command "$WIPEFS_PROG" wipefs
> +
> +# peal a (last) device from dev pool to be use it as replace target device
> +DEV_REPLACE_TGT="`echo $SCRATCH_DEV_POOL | $AWK_PROG '{print $NF}'`"
> +SCRATCH_DEV_POOL=`echo $SCRATCH_DEV_POOL | sed -e "s# *$DEV_REPLACE_TGT *##"`
> +
> +# create a dmerror device
> +# peal the last device in SCRATCH_DEV_POOL and assign it to DM_ERROR_DEV
> +DEV_REPLACE_SRC="`echo $SCRATCH_DEV_POOL | $AWK_PROG '{print $NF}'`"
> +SCRATCH_DEV_POOL=`echo $SCRATCH_DEV_POOL | sed -e "s# *$DEV_REPLACE_SRC *##"`
>
> +# Create dm linear device /dev/mapper/error-dev using dev_last
> +# DM_ERROR_DEV will have dm linear device set if created successfully
> +_require_dm_error $DEV_REPLACE_SRC
> +_init_dm_error_dev  $DEV_REPLACE_SRC
> +
> +# assign created dm device to the scratch pool
> +SCRATCH_DEV_POOL="$SCRATCH_DEV_POOL $DM_ERROR_DEV"

This is a lot of screwing around to modify the device list.

Just make a local copy of the device pool, clear the
SCRATCH_DEV_POOL variable, and pass the local devices to
_scratch_pool_mkfs....

> +
> +echo DEV_REPLACE_SRC=$DEV_REPLACE_SRC >> $seqres.full
> +echo DEV_REPLACE_TGT=$DEV_REPLACE_TGT >> $seqres.full
> +echo DM_ERROR_DEV=$DM_ERROR_DEV >> $seqres.full
> +echo SCRATCH_DEV_POOL=$SCRATCH_DEV_POOL >> $seqres.full
> +
> +MKFS_OPTIONS="-m raid1 -d raid1"
> +_scratch_pool_mkfs >> $seqres.full 2>&1 || _fail "mkfs failed"
> +_scratch_mount
> +
> +$BTRFS_UTIL_PROG filesystem show -m $SCRATCH_MNT >> $seqres.full
> +
> +error_dev_devid=`$BTRFS_UTIL_PROG filesystem show -m $SCRATCH_MNT |\
> +			egrep $DM_ERROR_DEV | $AWK_PROG '{print $2}'`
> +echo error_dev_devid=$error_dev_devid >> $seqres.full
> +
> +# now load the error into the DM_ERROR_DEV
> +_load_dm_error_table
> +
> +$BTRFS_UTIL_PROG replace start -B $error_dev_devid $DEV_REPLACE_TGT $SCRATCH_MNT -f ||\
> +		_fail "btrfs device replace failed, devid $error_dev_devid"

You haven't actually done any filesystem operations to trigger a
device error and have the filesystem react to it. before replacing
the device, you need to check that the filesystem doesn't fall over
on error, and that it recovers correctly once the device is
replaced.

> +echo "Silence is golden"

This test should not be silent - A user is running a maintenance
operation, and btrfs needs to be emitting a success of failure
message to the user.  The golden output should capture that
message, not use return values to determine if the test failed or
not.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com