| Message ID | 1382120790-31060-3-git-send-email-jbacik@fusionio.com (mailing list archive) |
|---|---|
| State | Not Applicable |
| Headers | show |
On 10/18/13 1:26 PM, Josef Bacik wrote: > I've been periodically failing btrfs/003 because my box sometimes takes a little > longer to unregister the device when we remove it and so the output from btrfs > dev show doesn't match what we are wanting since it still sees the device. To > fix this just stat and sleep if we still see the device node and only continue > once udev or whatever actually removes the device node so that we don't get > random failures. Thanks, > > Signed-off-by: Josef Bacik <jbacik@fusionio.com> > --- > V1->V2: Take Eric's suggestion to do this in the helper function > > common/rc | 9 +++++++++ > tests/btrfs/003 | 2 +- > 2 files changed, 10 insertions(+), 1 deletion(-) > > diff --git a/common/rc b/common/rc > index b253948..253bd05 100644 > --- a/common/rc > +++ b/common/rc > @@ -2093,7 +2093,16 @@ _require_freeze() > # ls -l /sys/class/block/sdd | rev | cut -d "/" -f 3 | rev > _devmgt_remove() > { > + local h=$1 $h is never used? You use ${1} directly below. Is that intentional? (Not sure if you just meant to eat $1 or if you forgot to replace it in the echo string). (if you respin, maybe give "$h" a more meaningful name?) -Eric > + local disk=$2 > + > echo 1 > /sys/class/scsi_device/${1}/device/delete || _fail "Remove disk failed" > + > + stat $disk > /dev/null 2>&1 > + while [ $? -eq 0 ]; do > + sleep 1 > + stat $disk > /dev/null 2>&1 > + done > } > > # arg 1 is dev to add and is output of the below eg. > diff --git a/tests/btrfs/003 b/tests/btrfs/003 > index 262b1d5..15c2cc7 100755 > --- a/tests/btrfs/003 > +++ b/tests/btrfs/003 > @@ -142,7 +142,7 @@ _test_replace() > DEVHTL=`ls -l /sys/class/block/${d} | rev | cut -d "/" -f 3 | rev` > > #fail disk > - _devmgt_remove ${DEVHTL} > + _devmgt_remove ${DEVHTL} $ds > dev_removed=1 > > $BTRFS_UTIL_PROG fi show $SCRATCH_DEV | grep "Some devices missing" >> $seqres.full || _fail \ > -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Oct 21, 2013 at 10:07:21AM -0500, Eric Sandeen wrote: > On 10/18/13 1:26 PM, Josef Bacik wrote: > > I've been periodically failing btrfs/003 because my box sometimes takes a little > > longer to unregister the device when we remove it and so the output from btrfs > > dev show doesn't match what we are wanting since it still sees the device. To > > fix this just stat and sleep if we still see the device node and only continue > > once udev or whatever actually removes the device node so that we don't get > > random failures. Thanks, > > > > Signed-off-by: Josef Bacik <jbacik@fusionio.com> > > --- > > V1->V2: Take Eric's suggestion to do this in the helper function > > > > common/rc | 9 +++++++++ > > tests/btrfs/003 | 2 +- > > 2 files changed, 10 insertions(+), 1 deletion(-) > > > > diff --git a/common/rc b/common/rc > > index b253948..253bd05 100644 > > --- a/common/rc > > +++ b/common/rc > > @@ -2093,7 +2093,16 @@ _require_freeze() > > # ls -l /sys/class/block/sdd | rev | cut -d "/" -f 3 | rev > > _devmgt_remove() > > { > > + local h=$1 > > $h is never used? You use ${1} directly below. Is that intentional? > (Not sure if you just meant to eat $1 or if you forgot to replace it > in the echo string). > > (if you respin, maybe give "$h" a more meaningful name?) > Yeah I screwed up, I'll respin. Thanks, Josef -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/common/rc b/common/rc index b253948..253bd05 100644 --- a/common/rc +++ b/common/rc @@ -2093,7 +2093,16 @@ _require_freeze() # ls -l /sys/class/block/sdd | rev | cut -d "/" -f 3 | rev _devmgt_remove() { + local h=$1 + local disk=$2 + echo 1 > /sys/class/scsi_device/${1}/device/delete || _fail "Remove disk failed" + + stat $disk > /dev/null 2>&1 + while [ $? -eq 0 ]; do + sleep 1 + stat $disk > /dev/null 2>&1 + done } # arg 1 is dev to add and is output of the below eg. diff --git a/tests/btrfs/003 b/tests/btrfs/003 index 262b1d5..15c2cc7 100755 --- a/tests/btrfs/003 +++ b/tests/btrfs/003 @@ -142,7 +142,7 @@ _test_replace() DEVHTL=`ls -l /sys/class/block/${d} | rev | cut -d "/" -f 3 | rev` #fail disk - _devmgt_remove ${DEVHTL} + _devmgt_remove ${DEVHTL} $ds dev_removed=1 $BTRFS_UTIL_PROG fi show $SCRATCH_DEV | grep "Some devices missing" >> $seqres.full || _fail \
I've been periodically failing btrfs/003 because my box sometimes takes a little longer to unregister the device when we remove it and so the output from btrfs dev show doesn't match what we are wanting since it still sees the device. To fix this just stat and sleep if we still see the device node and only continue once udev or whatever actually removes the device node so that we don't get random failures. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com> --- V1->V2: Take Eric's suggestion to do this in the helper function common/rc | 9 +++++++++ tests/btrfs/003 | 2 +- 2 files changed, 10 insertions(+), 1 deletion(-)