Skip to content

Commit

Permalink
rgw: add rgw-restore-bucket-index
Browse files Browse the repository at this point in the history
Adds an experimental script that allows a bucket index of a
non-versioned bucket to be restored by applying `radosgw-admin object
reindex ...` to all objects in the specified bucket. The objects in
the bucket are determined by scanning the data pool for head objects
containing the bucket's marker.

Signed-off-by: J. Eric Ivancich <[email protected]>
  • Loading branch information
ivancich committed Mar 9, 2023
1 parent be0583c commit 6ea11b1
Show file tree
Hide file tree
Showing 4 changed files with 196 additions and 0 deletions.
1 change: 1 addition & 0 deletions ceph.spec.in
Original file line number Diff line number Diff line change
Expand Up @@ -2132,6 +2132,7 @@ fi
%{_bindir}/rgw-gap-list-comparator
%{_bindir}/rgw-orphan-list
%{_bindir}/rgw-policy-check
%{_bindir}/rgw-restore-bucket-index
%{_mandir}/man8/radosgw.8*
%{_mandir}/man8/rgw-policy-check.8*
%dir %{_localstatedir}/lib/ceph/radosgw
Expand Down
1 change: 1 addition & 0 deletions debian/radosgw.install
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@ usr/bin/radosgw-token
usr/bin/rgw-gap-list
usr/bin/rgw-gap-list-comparator
usr/bin/rgw-orphan-list
usr/bin/rgw-restore-bucket-index
usr/share/man/man8/ceph-diff-sorted.8
usr/share/man/man8/radosgw.8
usr/share/man/man8/rgw-orphan-list.8
1 change: 1 addition & 0 deletions src/rgw/CMakeLists.txt
Original file line number Diff line number Diff line change
Expand Up @@ -578,4 +578,5 @@ install(PROGRAMS
rgw-gap-list
rgw-gap-list-comparator
rgw-orphan-list
rgw-restore-bucket-index
DESTINATION bin)
193 changes: 193 additions & 0 deletions src/rgw/rgw-restore-bucket-index
Original file line number Diff line number Diff line change
@@ -0,0 +1,193 @@
#!/usr/bin/env bash

# version 2023-03-07

# rgw-restore-bucket-index is an EXPERIMENTAL tool to use in case
# bucket index entries for objects in the bucket are somehow lost. It
# is expected to be needed and used rarely. A bucket name is provided
# and the data pool for that bucket is scanned for all head objects
# matching the bucket's marker. The rgw object name is then extracted
# from the rados object name, and `radosgw-admin bucket reindex ...`
# is used to add the bucket index entry.
#
# Because this script must process json objects, the `jq` tool must be
# installed on the system.
#
# Usage: $0 [--proceed] <bucket-name> [data-pool-name]
#
# This tool is designed to be interactive, allowing the user to
# examine the list of objects to be reindexed before
# proceeding. However, if the "--proceed" option is provided, the
# script will not prompt the user and simply proceed.

trap "clean ; exit 1" TERM
export TOP_PID=$$

# IMPORTANT: affects order produced by 'sort' and 'ceph-diff-sorted'
# relies on this ordering
export LC_ALL=C

export bkt_entry=/tmp/rgwrbi-bkt-entry.$$
export bkt_inst=/tmp/rgwrbi-bkt-inst.$$
export bkt_inst_new=/tmp/rgwrbi-bkt-inst-new.$$
export obj_list=/tmp/rgwrbi-object-list.$$
export zone_info=/tmp/rgwrbi-zone-info.$$
export clean_temps=1

# make sure jq is available
if which jq > /dev/null ;then
:
else
echo 'Error: must have command `jq` installed and on $PATH for json parsing.'
exit 1
fi

clean() {
if [ -n "$clean_temps" ] ;then
rm -f $bkt_entry $bkt_inst $bkt_inst_new $obj_list $zone_info
fi
}

super_exit() {
kill -s TERM $TOP_PID
}

usage() {
>&2 cat << EOF
Usage: $0 [--proceed] <bucket-name> [data-pool-name]
NOTE: This tool is currently considered EXPERIMENTAL.
NOTE: If a data-pool-name is not supplied then it will be inferred from bucket and zone information.
NOTE: If --proceed is provided then user will not be prompted to proceed. Use with caution.
EOF
super_exit
}

# strips the starting and ending double quotes from a string, so:
# "dog" -> dog
# "dog -> "dog
# d"o"g -> d"o"g
# "do"g" -> do"g
strip_quotes() {
echo "$1" | sed 's/^"\(.*\)"$/\1/'
}

# Determines the name of the data pool. Expects the optional
# command-line argument to appear as $1 if there is one. The
# command-line has the highest priority, then the "explicit_placement"
# in the bucket instance data, and finally the "placement_rule" in the
# bucket instance data.
get_pool() {
# command-line
if [ -n "$1" ] ;then
echo "$1"
exit 0
fi

# explicit_placement
expl_pool=$(strip_quotes $(jq '.data.bucket_info.bucket.explicit_placement.data_pool' $bkt_inst))
if [ -n "$expl_pool" ] ;then
echo "$expl_pool"
exit 0
fi

# placement_rule
plmt_rule=$(strip_quotes $(jq '.data.bucket_info.placement_rule' $bkt_inst))
plmt_pool=$(echo "$plmt_rule" | awk -F / '{print $1}')
plmt_class=$(echo "$plmt_rule" | awk -F / '{print $2}')
if [ -z "$plmt_class" ] ;then
plmt_class=STANDARD
fi

radosgw-admin zone get >$zone_info 2>/dev/null
pool=$(strip_quotes $(jq ".placement_pools [] | select(.key | contains(\"${plmt_pool}\")) .val .storage_classes.${plmt_class}.data_pool" $zone_info))

if [ -z "$pool" ] ;then
echo ERROR: unable to determine pool.
super_exit
fi
echo "$pool"
}

if [ $1 == "--proceed" ] ;then
echo "NOTICE: This tool is currently considered EXPERIMENTAL."
proceed=1
shift
fi

# expect 1 or 2 arguments
if [ $# -eq 0 -o $# -gt 2 ] ;then
usage
fi

bucket=$1

# read bucket entry metadata
radosgw-admin metadata get bucket:$bucket >$bkt_entry 2>/dev/null
marker=$(strip_quotes $(jq ".data.bucket.marker" $bkt_entry))
bucket_id=$(strip_quotes $(jq ".data.bucket.bucket_id" $bkt_entry))
echo marker is $marker
echo bucket_id is $bucket_id

# read bucket instance metadata
radosgw-admin metadata get bucket.instance:${bucket}:$bucket_id >$bkt_inst 2>/dev/null

# handle versioned buckets
bkt_flags=$(jq ".data.bucket_info.flags" $bkt_inst)
is_versioned=$(( $bkt_flags & 2)) # mask bit indicating it's a versioned bucket
if [ "$is_versioned" -ne 0 ] ;then
echo "Error: this bucket appears to be versioned, and this tool cannot work with versioned buckets."
clean
exit 1
fi

# examine number of bucket index shards
num_shards=$(jq ".data.bucket_info.num_shards" $bkt_inst)
echo number of bucket index shards is $num_shards

# determine data pool
pool=$(get_pool $2)
echo data pool is $pool

# search the data pool for all of the head objects that begin with the
# marker that are not in namespaces (indicated by an extra underscore)
# and then strip away all but the rgw object name
( rados -p $pool ls | grep "^${marker}_[^_]" | sed "s/^${marker}_\(.*\)/\1/" >$obj_list ) 2>/dev/null

# handle the case where the resulting object list file is empty
if [ -s $obj_list ] ;then
:
else
echo "NOTICE: No head objects for bucket \"$bucket\" were found in pool \"$pool\", so nothing was recovered."
clean
exit 0
fi

if [ -z "$proceed" ] ;then
# warn user and get permission to proceed
echo "NOTICE: This tool is currently considered EXPERIMENTAL."
echo "The list of objects that we will attempt to restore can be found in \"$obj_list\"."
echo "Please review the object names in that file (either below or in another window/terminal) before proceeding."
while true ; do
read -p "Type \"proceed!\" to proceed, \"view\" to view object list, or \"q\" to quit: " action
if [ "$action" == "q" ] ;then
echo "Exiting..."
clean
exit 0
elif [ "$action" == "view" ] ;then
echo "Viewing..."
less $obj_list
elif [ "$action" == "proceed!" ] ;then
echo "Proceeding..."
break
else
echo "Error: response \"$action\" is not understood."
fi
done
fi

# execute object rewrite on all of the head objects
radosgw-admin object reindex --bucket=$bucket --objects-file=$obj_list 2>/dev/null

clean
echo Done

0 comments on commit 6ea11b1

Please sign in to comment.