openstack-helm-infra/ceph-osd/templates/bin/osd/ceph-volume/_bluestore.sh.tpl
Stephen Taylor 885285139e [ceph-osd] Alias synchronized commands and fix descriptor leak
There are many race conditions possible when multiple ceph-osd
pods are initialized on the same host at the same time using
shared metadata disks. The locked() function was introduced a
while back to address these, but some commands weren't locked,
locked() was being called all over the place, and there was a file
descriptor leak in locked(). This change cleans that up by
by maintaining a single, global file descriptor for the lock file
that is only opened and closed once, and also by aliasing all of
the commands that need to use locked() and removing explicit calls
to locked() everywhere.

The global_locked() function has also been removed as it isn't
needed when individual commands that interact with disks use
locked() properly.

Change-Id: I0018cf0b3a25bced44c57c40e33043579c42de7a
2020-12-16 07:22:15 -07:00

117 lines
3.3 KiB
Smarty

#!/bin/bash
{{/*
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/}}
source /tmp/osd-common-ceph-volume.sh
set -ex
: "${OSD_SOFT_FORCE_ZAP:=1}"
export OSD_DEVICE=$(readlink -f ${STORAGE_LOCATION})
if [[ -z "${OSD_DEVICE}" ]];then
echo "ERROR- You must provide a device to build your OSD ie: /dev/sdb"
exit 1
fi
if [[ ! -b "${OSD_DEVICE}" ]]; then
echo "ERROR- The device pointed by OSD_DEVICE ${OSD_DEVICE} doesn't exist !"
exit 1
fi
CEPH_DISK_OPTIONS=""
CEPH_OSD_OPTIONS=""
udev_settle
OSD_ID=$(get_osd_id_from_device ${OSD_DEVICE})
simple_activate=0
if [[ -z ${OSD_ID} ]]; then
echo "Looks like ceph-disk has been used earlier to activate the OSD."
tmpmnt=$(mktemp -d)
mount ${OSD_DEVICE}1 ${tmpmnt}
OSD_ID=$(cat ${tmpmnt}/whoami)
umount ${tmpmnt}
simple_activate=1
fi
OSD_FSID=$(get_osd_fsid_from_device ${OSD_DEVICE})
if [[ -z ${OSD_FSID} ]]; then
echo "Looks like ceph-disk has been used earlier to activate the OSD."
tmpmnt=$(mktemp -d)
mount ${OSD_DEVICE}1 ${tmpmnt}
OSD_FSID=$(cat ${tmpmnt}/fsid)
umount ${tmpmnt}
simple_activate=1
fi
OSD_PATH="${OSD_PATH_BASE}-${OSD_ID}"
OSD_KEYRING="${OSD_PATH}/keyring"
mkdir -p ${OSD_PATH}
if [[ ${simple_activate} -eq 1 ]]; then
ceph-volume simple activate --no-systemd ${OSD_ID} ${OSD_FSID}
else
ceph-volume lvm -v \
--setuser ceph \
--setgroup disk \
activate ${CEPH_DISK_OPTIONS} \
--auto-detect-objectstore \
--no-systemd ${OSD_ID} ${OSD_FSID}
# Cross check the db and wal symlinks if missed
DB_DEV=$(get_osd_db_device_from_device ${OSD_DEVICE})
if [[ ! -z ${DB_DEV} ]]; then
if [[ ! -h /var/lib/ceph/osd/ceph-${OSD_ID}/block.db ]]; then
ln -snf ${DB_DEV} /var/lib/ceph/osd/ceph-${OSD_ID}/block.db
chown -h ceph:ceph ${DB_DEV}
chown -h ceph:ceph /var/lib/ceph/osd/ceph-${OSD_ID}/block.db
fi
fi
WAL_DEV=$(get_osd_wal_device_from_device ${OSD_DEVICE})
if [[ ! -z ${WAL_DEV} ]]; then
if [[ ! -h /var/lib/ceph/osd/ceph-${OSD_ID}/block.wal ]]; then
ln -snf ${WAL_DEV} /var/lib/ceph/osd/ceph-${OSD_ID}/block.wal
chown -h ceph:ceph ${WAL_DEV}
chown -h ceph:ceph /var/lib/ceph/osd/ceph-${OSD_ID}/block.wal
fi
fi
fi
# NOTE(stevetaylor): Set the OSD's crush weight (use noin flag to prevent rebalancing if necessary)
OSD_WEIGHT=$(get_osd_crush_weight_from_device ${OSD_DEVICE})
# NOTE(supamatt): add or move the OSD's CRUSH location
crush_location
# NOTE(supamatt): Just in case permissions do not align up, we recursively set them correctly.
if [ $(stat -c%U ${OSD_PATH}) != ceph ]; then
chown -R ceph. ${OSD_PATH};
fi
# NOTE(gagehugo): Writing the OSD_ID to tmp for logging
echo "${OSD_ID}" > /tmp/osd-id
exec /usr/bin/ceph-osd \
--cluster ${CLUSTER} \
${CEPH_OSD_OPTIONS} \
-f \
-i ${OSD_ID} \
--setuser ceph \
--setgroup disk & echo $! > /run/ceph-osd.pid
wait
# Clean up resources held by the common script
common_cleanup