frostfs-node/pkg/local_object_storage/metabase/select.go

563 lines
14 KiB
Go
Raw Permalink Normal View History

package meta
import (
"encoding/binary"
"errors"
"fmt"
"strings"
v2object "github.com/TrueCloudLab/frostfs-api-go/v2/object"
cid "github.com/TrueCloudLab/frostfs-sdk-go/container/id"
"github.com/TrueCloudLab/frostfs-sdk-go/object"
oid "github.com/TrueCloudLab/frostfs-sdk-go/object/id"
"go.etcd.io/bbolt"
"go.uber.org/zap"
)
type (
// filterGroup is a structure that have search filters grouped by access
// method. We have fast filters that looks for indexes and do not unmarshal
// objects, and slow filters, that applied after fast filters created
// smaller set of objects to check.
filterGroup struct {
withCnrFilter bool
cnr cid.ID
fastFilters, slowFilters object.SearchFilters
}
)
// SelectPrm groups the parameters of Select operation.
type SelectPrm struct {
cnr cid.ID
filters object.SearchFilters
}
// SelectRes groups the resulting values of Select operation.
type SelectRes struct {
addrList []oid.Address
}
// SetContainerID is a Select option to set the container id to search in.
func (p *SelectPrm) SetContainerID(cnr cid.ID) {
p.cnr = cnr
}
// SetFilters is a Select option to set the object filters.
func (p *SelectPrm) SetFilters(fs object.SearchFilters) {
p.filters = fs
}
// AddressList returns list of addresses of the selected objects.
func (r SelectRes) AddressList() []oid.Address {
return r.addrList
}
// Select returns list of addresses of objects that match search filters.
func (db *DB) Select(prm SelectPrm) (res SelectRes, err error) {
db.modeMtx.RLock()
defer db.modeMtx.RUnlock()
if db.mode.NoMetabase() {
return res, ErrDegradedMode
}
if blindlyProcess(prm.filters) {
return res, nil
}
currEpoch := db.epochState.CurrentEpoch()
return res, db.boltDB.View(func(tx *bbolt.Tx) error {
res.addrList, err = db.selectObjects(tx, prm.cnr, prm.filters, currEpoch)
return err
})
}
func (db *DB) selectObjects(tx *bbolt.Tx, cnr cid.ID, fs object.SearchFilters, currEpoch uint64) ([]oid.Address, error) {
group, err := groupFilters(fs)
if err != nil {
return nil, err
}
// if there are conflicts in query and container then it means that there is no
// objects to match this query.
if group.withCnrFilter && !cnr.Equals(group.cnr) {
return nil, nil
}
// keep matched addresses in this cache
// value equal to number (index+1) of latest matched filter
mAddr := make(map[string]int)
expLen := len(group.fastFilters) // expected value of matched filters in mAddr
if len(group.fastFilters) == 0 {
expLen = 1
db.selectAll(tx, cnr, mAddr)
} else {
for i := range group.fastFilters {
db.selectFastFilter(tx, cnr, group.fastFilters[i], mAddr, i)
}
}
res := make([]oid.Address, 0, len(mAddr))
for a, ind := range mAddr {
if ind != expLen {
continue // ignore objects with unmatched fast filters
}
var id oid.ID
err = id.Decode([]byte(a))
if err != nil {
return nil, err
}
var addr oid.Address
addr.SetContainer(cnr)
addr.SetObject(id)
if objectStatus(tx, addr, currEpoch) > 0 {
continue // ignore removed objects
}
if !db.matchSlowFilters(tx, addr, group.slowFilters, currEpoch) {
continue // ignore objects with unmatched slow filters
}
res = append(res, addr)
}
return res, nil
}
// selectAll adds to resulting cache all available objects in metabase.
func (db *DB) selectAll(tx *bbolt.Tx, cnr cid.ID, to map[string]int) {
bucketName := make([]byte, bucketKeySize)
selectAllFromBucket(tx, primaryBucketName(cnr, bucketName), to, 0)
selectAllFromBucket(tx, tombstoneBucketName(cnr, bucketName), to, 0)
selectAllFromBucket(tx, storageGroupBucketName(cnr, bucketName), to, 0)
selectAllFromBucket(tx, parentBucketName(cnr, bucketName), to, 0)
selectAllFromBucket(tx, bucketNameLockers(cnr, bucketName), to, 0)
}
// selectAllFromBucket goes through all keys in bucket and adds them in a
// resulting cache. Keys should be stringed object ids.
func selectAllFromBucket(tx *bbolt.Tx, name []byte, to map[string]int, fNum int) {
bkt := tx.Bucket(name)
if bkt == nil {
return
}
_ = bkt.ForEach(func(k, v []byte) error {
markAddressInCache(to, fNum, string(k))
return nil
})
}
// selectFastFilter makes fast optimized checks for well known buckets or
// looking through user attribute buckets otherwise.
func (db *DB) selectFastFilter(
tx *bbolt.Tx,
cnr cid.ID, // container we search on
f object.SearchFilter, // fast filter
to map[string]int, // resulting cache
fNum int, // index of filter
) {
currEpoch := db.epochState.CurrentEpoch()
bucketName := make([]byte, bucketKeySize)
switch f.Header() {
case v2object.FilterHeaderObjectID:
db.selectObjectID(tx, f, cnr, to, fNum, currEpoch)
case v2object.FilterHeaderOwnerID:
bucketName := ownerBucketName(cnr, bucketName)
db.selectFromFKBT(tx, bucketName, f, to, fNum)
case v2object.FilterHeaderPayloadHash:
bucketName := payloadHashBucketName(cnr, bucketName)
db.selectFromList(tx, bucketName, f, to, fNum)
case v2object.FilterHeaderObjectType:
for _, bucketName := range bucketNamesForType(cnr, f.Operation(), f.Value()) {
selectAllFromBucket(tx, bucketName, to, fNum)
}
case v2object.FilterHeaderParent:
bucketName := parentBucketName(cnr, bucketName)
db.selectFromList(tx, bucketName, f, to, fNum)
case v2object.FilterHeaderSplitID:
bucketName := splitBucketName(cnr, bucketName)
db.selectFromList(tx, bucketName, f, to, fNum)
case v2object.FilterPropertyRoot:
selectAllFromBucket(tx, rootBucketName(cnr, bucketName), to, fNum)
case v2object.FilterPropertyPhy:
selectAllFromBucket(tx, primaryBucketName(cnr, bucketName), to, fNum)
selectAllFromBucket(tx, tombstoneBucketName(cnr, bucketName), to, fNum)
selectAllFromBucket(tx, storageGroupBucketName(cnr, bucketName), to, fNum)
selectAllFromBucket(tx, bucketNameLockers(cnr, bucketName), to, fNum)
default: // user attribute
bucketName := attributeBucketName(cnr, f.Header(), bucketName)
if f.Operation() == object.MatchNotPresent {
selectOutsideFKBT(tx, allBucketNames(cnr), bucketName, to, fNum)
} else {
db.selectFromFKBT(tx, bucketName, f, to, fNum)
}
}
}
var mBucketNaming = map[string][]func(cid.ID, []byte) []byte{
v2object.TypeRegular.String(): {primaryBucketName, parentBucketName},
v2object.TypeTombstone.String(): {tombstoneBucketName},
v2object.TypeStorageGroup.String(): {storageGroupBucketName},
v2object.TypeLock.String(): {bucketNameLockers},
}
func allBucketNames(cnr cid.ID) (names [][]byte) {
for _, fns := range mBucketNaming {
for _, fn := range fns {
names = append(names, fn(cnr, make([]byte, bucketKeySize)))
}
}
return
}
func bucketNamesForType(cnr cid.ID, mType object.SearchMatchType, typeVal string) (names [][]byte) {
appendNames := func(key string) {
fns, ok := mBucketNaming[key]
if ok {
for _, fn := range fns {
names = append(names, fn(cnr, make([]byte, bucketKeySize)))
}
}
}
switch mType {
default:
case object.MatchStringNotEqual:
for key := range mBucketNaming {
if key != typeVal {
appendNames(key)
}
}
case object.MatchStringEqual:
appendNames(typeVal)
case object.MatchCommonPrefix:
for key := range mBucketNaming {
if strings.HasPrefix(key, typeVal) {
appendNames(key)
}
}
}
return
}
// selectFromList looks into <fkbt> index to find list of addresses to add in
// resulting cache.
func (db *DB) selectFromFKBT(
tx *bbolt.Tx,
name []byte, // fkbt root bucket name
f object.SearchFilter, // filter for operation and value
to map[string]int, // resulting cache
fNum int, // index of filter
) { //
matchFunc, ok := db.matchers[f.Operation()]
if !ok {
db.log.Debug("missing matcher", zap.Uint32("operation", uint32(f.Operation())))
return
}
fkbtRoot := tx.Bucket(name)
if fkbtRoot == nil {
return
}
err := matchFunc.matchBucket(fkbtRoot, f.Header(), f.Value(), func(k, _ []byte) error {
fkbtLeaf := fkbtRoot.Bucket(k)
if fkbtLeaf == nil {
return nil
}
return fkbtLeaf.ForEach(func(k, _ []byte) error {
markAddressInCache(to, fNum, string(k))
return nil
})
})
if err != nil {
db.log.Debug("error in FKBT selection", zap.String("error", err.Error()))
}
}
// selectOutsideFKBT looks into all incl buckets to find list of addresses outside <fkbt> to add in
// resulting cache.
func selectOutsideFKBT(
tx *bbolt.Tx,
incl [][]byte, // buckets
name []byte, // fkbt root bucket name
to map[string]int, // resulting cache
fNum int, // index of filter
) {
mExcl := make(map[string]struct{})
bktExcl := tx.Bucket(name)
if bktExcl != nil {
_ = bktExcl.ForEach(func(k, _ []byte) error {
exclBktLeaf := bktExcl.Bucket(k)
if exclBktLeaf == nil {
return nil
}
return exclBktLeaf.ForEach(func(k, _ []byte) error {
mExcl[string(k)] = struct{}{}
return nil
})
})
}
for i := range incl {
bktIncl := tx.Bucket(incl[i])
if bktIncl == nil {
continue
}
_ = bktIncl.ForEach(func(k, _ []byte) error {
if _, ok := mExcl[string(k)]; !ok {
markAddressInCache(to, fNum, string(k))
}
return nil
})
}
}
// selectFromList looks into <list> index to find list of addresses to add in
// resulting cache.
func (db *DB) selectFromList(
tx *bbolt.Tx,
name []byte, // list root bucket name
f object.SearchFilter, // filter for operation and value
to map[string]int, // resulting cache
fNum int, // index of filter
) { //
bkt := tx.Bucket(name)
if bkt == nil {
return
}
var (
lst [][]byte
err error
)
switch op := f.Operation(); op {
case object.MatchStringEqual:
lst, err = decodeList(bkt.Get(bucketKeyHelper(f.Header(), f.Value())))
if err != nil {
db.log.Debug("can't decode list bucket leaf", zap.String("error", err.Error()))
return
}
default:
fMatch, ok := db.matchers[op]
if !ok {
db.log.Debug("unknown operation", zap.Uint32("operation", uint32(op)))
return
}
if err = fMatch.matchBucket(bkt, f.Header(), f.Value(), func(key, val []byte) error {
l, err := decodeList(val)
if err != nil {
db.log.Debug("can't decode list bucket leaf",
zap.String("error", err.Error()),
)
return err
}
lst = append(lst, l...)
return nil
}); err != nil {
db.log.Debug("can't iterate over the bucket",
zap.String("error", err.Error()),
)
return
}
}
for i := range lst {
markAddressInCache(to, fNum, string(lst[i]))
}
}
// selectObjectID processes objectID filter with in-place optimizations.
func (db *DB) selectObjectID(
tx *bbolt.Tx,
f object.SearchFilter,
cnr cid.ID,
to map[string]int, // resulting cache
fNum int, // index of filter
currEpoch uint64,
) {
appendOID := func(id oid.ID) {
var addr oid.Address
addr.SetContainer(cnr)
addr.SetObject(id)
ok, err := db.exists(tx, addr, currEpoch)
if (err == nil && ok) || errors.As(err, &splitInfoError) {
raw := make([]byte, objectKeySize)
id.Encode(raw)
markAddressInCache(to, fNum, string(raw))
}
}
switch op := f.Operation(); op {
case object.MatchStringEqual:
var id oid.ID
if err := id.DecodeString(f.Value()); err == nil {
appendOID(id)
}
default:
fMatch, ok := db.matchers[op]
if !ok {
db.log.Debug("unknown operation",
zap.Uint32("operation", uint32(f.Operation())),
)
return
}
for _, bucketName := range bucketNamesForType(cnr, object.MatchStringNotEqual, "") {
// copy-paste from DB.selectAllFrom
bkt := tx.Bucket(bucketName)
if bkt == nil {
return
}
err := fMatch.matchBucket(bkt, f.Header(), f.Value(), func(k, v []byte) error {
var id oid.ID
if err := id.Decode(k); err == nil {
appendOID(id)
}
return nil
})
if err != nil {
db.log.Debug("could not iterate over the buckets",
zap.String("error", err.Error()),
)
}
}
}
}
// matchSlowFilters return true if object header is matched by all slow filters.
func (db *DB) matchSlowFilters(tx *bbolt.Tx, addr oid.Address, f object.SearchFilters, currEpoch uint64) bool {
if len(f) == 0 {
return true
}
buf := make([]byte, addressKeySize)
obj, err := db.get(tx, addr, buf, true, false, currEpoch)
if err != nil {
return false
}
for i := range f {
matchFunc, ok := db.matchers[f[i].Operation()]
if !ok {
return false
}
var data []byte
switch f[i].Header() {
case v2object.FilterHeaderVersion:
data = []byte(obj.Version().String())
case v2object.FilterHeaderHomomorphicHash:
cs, _ := obj.PayloadHomomorphicHash()
data = cs.Value()
case v2object.FilterHeaderCreationEpoch:
data = make([]byte, 8)
binary.LittleEndian.PutUint64(data, obj.CreationEpoch())
case v2object.FilterHeaderPayloadLength:
data = make([]byte, 8)
binary.LittleEndian.PutUint64(data, obj.PayloadSize())
default:
continue // ignore unknown search attributes
}
if !matchFunc.matchSlow(f[i].Header(), data, f[i].Value()) {
return false
}
}
return true
}
// groupFilters divides filters in two groups: fast and slow. Fast filters
// processed by indexes and slow filters processed after by unmarshaling
// object headers.
func groupFilters(filters object.SearchFilters) (filterGroup, error) {
res := filterGroup{
fastFilters: make(object.SearchFilters, 0, len(filters)),
slowFilters: make(object.SearchFilters, 0, len(filters)),
}
for i := range filters {
switch filters[i].Header() {
case v2object.FilterHeaderContainerID: // support deprecated field
err := res.cnr.DecodeString(filters[i].Value())
if err != nil {
return filterGroup{}, fmt.Errorf("can't parse container id: %w", err)
}
res.withCnrFilter = true
case // slow filters
v2object.FilterHeaderVersion,
v2object.FilterHeaderCreationEpoch,
v2object.FilterHeaderPayloadLength,
v2object.FilterHeaderHomomorphicHash:
res.slowFilters = append(res.slowFilters, filters[i])
default: // fast filters or user attributes if unknown
res.fastFilters = append(res.fastFilters, filters[i])
}
}
return res, nil
}
func markAddressInCache(cache map[string]int, fNum int, addr string) {
if num := cache[addr]; num == fNum {
cache[addr] = num + 1
}
}
// returns true if query leads to a deliberately empty result.
func blindlyProcess(fs object.SearchFilters) bool {
for i := range fs {
if fs[i].Operation() == object.MatchNotPresent && isSystemKey(fs[i].Header()) {
return true
}
// TODO: #1148 check other cases
// e.g. (a == b) && (a != b)
}
return false
}
// returns true if string key is a reserved system filter key.
func isSystemKey(key string) bool {
// FIXME: #1147 version-dependent approach
return strings.HasPrefix(key, v2object.ReservedFilterPrefix)
}