mirror of
https://github.com/seaweedfs/seaweedfs.git
synced 2026-09-14 02:20:41 +02:00
* feat(s3api): apply lifecycle TTL at write time The S3 server already has the bucket's lifecycle XML at PUT time (via the cached BucketConfig), so volume-TTL routing is just a per-write decision instead of something that needs a separate filer.conf projection kept in sync via operator commands. - BucketConfig caches the canonical Rules parsed from the lifecycle XML once on load (BucketConfigCache invalidates on Put/Delete Lifecycle, so the rules stay current automatically). - resolveLifecycleTTLForWrite walks the cached rules: longest-prefix match, applies tag and size filters against the request, returns Days * 86400. Versioned buckets, non-Expiration.Days rules, and unevaluable size filters (no Content-Length) yield 0 — the lifecycle worker handles those at scan time. - putToFiler resolves TTL once and passes it through both the AssignVolumeRequest (so chunks land on a TTL volume) and the new entry's Attributes.TtlSec (so the filer's RocksDB compaction also expires the metadata). Lifecycle XML PUT/DELETE now influences write routing immediately — no operator command, no filer.conf bookkeeping. The lifecycle worker remains authoritative for the cases the fast path can't cover (existing objects via bootstrap, versioned buckets, noncurrent retention, abort-MPU, tag/size filters that didn't hold at PUT time). CompleteMultipartUpload and CopyObject still need wiring; left for follow-ups so this PR stays scoped. * perf(s3api): pre-filter and sort lifecycle rules for the per-PUT TTL walk resolveLifecycleTTLForWrite walked every lifecycle rule on every PutObject, including disabled / non-Expiration.Days rules that could never fire on the fast path, and computed "longest prefix wins" via a running max instead of an early exit. Cache a pre-filtered + pre-sorted slice in BucketConfig: - buildTTLFastPathRules drops everything except Status=Enabled + ExpirationDays>0; - sorts by descending prefix length (stable, so equal-length rules keep their XML order). The resolver returns on first prefix+filter match. A bucket whose lifecycle XML has no Expiration.Days rules is now O(1); a typical bucket with one Expiration.Days rule walks one HasPrefix per PUT. The cache is built once per bucket-config load. PutBucketLifecycle / DeleteBucketLifecycle already invalidate the cache, so the fast-path slice stays current automatically. * refactor(s3api): LifecycleTTLResolver object + four review fixes Pulls the per-PUT TTL resolution into a dedicated type so the bucket config holds one object instead of a slice + magic-walk function: - LifecycleTTLResolver wraps the pre-filtered, pre-sorted rules. nil-safe Resolve so the call site doesn't have to special-case buckets with no eligible rules. Four review findings: 1. (high) drop tag-filtered rules from the fast path. Tags are mutable post-PUT via PutObjectTagging but volume TTL is irreversible — an object that matched at write time would still expire after the tag was removed. Worker re-evaluates current tags at scan time. Fast path now keeps only stable predicates: prefix and size. 2. (high) move TTL resolution out of putToFiler. MPU parts, copy-part destinations, and other transient writes called putToFiler with object="" — bucket-wide rules (empty Prefix) matched and bound a TTL clock starting at part-upload time, before CompleteMultipartUpload existed. putToFiler now takes an explicit ttlSec parameter; only the user-visible PutObject paths (PutObjectHandler, postpolicy) feed it from the resolver. MPU and copy-part pass 0. 3. (medium) AWS overlapping-rule precedence is "shorter expiration wins", not "longest prefix wins". Sort by ExpirationDays ascending so the first prefix match is also the shortest applicable rule. 4. (medium) overflow no longer caps at math.MaxInt32 seconds (~68y). A longer policy would have expired early. Return 0 instead so the worker enforces the actual policy on its own schedule. Versioning gate moves into the resolver constructor — versioned buckets get a nil resolver. The five putToFiler callers all updated: PutObjectHandler + postpolicy resolve via lifecycleTTLForObjectWrite, suspended/versioned wrappers pass 0 by construction, MPU part and copy-part SSE pass 0 with a one-line comment about why. * refactor(s3api): drop unused BucketConfig.LifecycleRules field The full canonical rule set was set on every bucket-config load but never read — resolveLifecycleTTLForWrite worked off the resolver's filtered slice, and the lifecycle worker reads bucket entries straight off the meta-log instead of this cache. Remove the field and its s3lifecycle import. * perf(s3api): pre-compute LifecycleTTLResolver hot-path fields Resolve was doing per-call work that's actually constant per bucket- config load: int64 multiplication, max-int32 overflow check, field indirections through *s3lifecycle.Rule. Move it to the constructor and pack the rule into a compact ttlRule (prefix + ttlSec int32 + sizeGT/sizeLT) so the inner loop is HasPrefix → optional size check → return. Drop overflowing rules at construction rather than handling per- resolve: capping would expire long policies early, and returning 0 in the inner loop would prevent any shorter overlapping rule from firing. Drop-at-construction composes correctly with the ascending sort. Benchmarks (Apple M4): NilReceiver 0.99 ns/op 0 B/op OneRuleMatching 2.75 ns/op 0 B/op FiveRulesNoMatch 13.5 ns/op 0 B/op * fix(s3api): refresh LifecycleTTL resolver on bucket-config update storeBucketLifecycleConfiguration writes to Entry.Extended via updateBucketConfig, which clones the cached BucketConfig and calls the user fn, then caches the result. The clone inherits the prior LifecycleTTL pointer and nothing rebuilt it from the new XML, so add/replace/delete of a lifecycle policy left the wrong resolver in cache until eviction. Same gap on the meta-log side: peer-driven updates flowed through updateBucketConfigCacheFromEntry without re-deriving the resolver. Centralize the Entry -> derived-field mapping in one helper that resets every Extended-backed field then repopulates from the entry, and call it from getBucketConfig (initial load), updateBucketConfig (after updateEntry succeeds, before caching), and updateBucketConfigCacheFromEntry (meta-log path). Reset is the load-bearing part: deleting the lifecycle XML must yield a nil resolver, since stamping a stale TTL onto subsequent writes is irreversible. * fix(s3api): PostPolicy passes object size, not multipart wire size lifecycleTTLForObjectWrite was reading r.ContentLength, which on the PostPolicy path is the multipart envelope (form fields + boundaries), not the uploaded object body. A size-filtered rule would evaluate against that inflated total and stamp (or skip) a TTL the policy didn't intend. Take the object size as an explicit parameter. PutObject still passes r.ContentLength (correct there); PostPolicy passes the fileSize already extracted from the form part. Negative size means unknown and continues to skip any size-filtered rule. * fix(s3api): treat Object Lock as versioned for lifecycle TTL fast path Object Lock requires versioning at the API level, but it can be enabled at create time without S3 ever writing the explicit Versioning header. The lifecycle resolver construction site only checked Versioning, so an Object-Lock bucket with no Versioning byte would still get a fast-path resolver and stamp volume TTL onto writes — destroying noncurrent versions when the volume expires. Mirror the OR already used in BucketIsVersioned: ObjectLockConfig non-nil counts as versioned for resolver construction. Existing explicit-Versioning paths are unchanged.
1199 lines
40 KiB
Go
1199 lines
40 KiB
Go
package s3api
|
|
|
|
import (
|
|
"context"
|
|
"encoding/json"
|
|
"errors"
|
|
"fmt"
|
|
"path/filepath"
|
|
"strings"
|
|
"sync"
|
|
"time"
|
|
|
|
"github.com/aws/aws-sdk-go/service/s3"
|
|
"google.golang.org/protobuf/proto"
|
|
|
|
"github.com/seaweedfs/seaweedfs/weed/glog"
|
|
"github.com/seaweedfs/seaweedfs/weed/kms"
|
|
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
|
|
"github.com/seaweedfs/seaweedfs/weed/pb/s3_pb"
|
|
"github.com/seaweedfs/seaweedfs/weed/s3api/cors"
|
|
"github.com/seaweedfs/seaweedfs/weed/s3api/lifecycle_xml"
|
|
"github.com/seaweedfs/seaweedfs/weed/s3api/policy_engine"
|
|
"github.com/seaweedfs/seaweedfs/weed/s3api/s3_constants"
|
|
"github.com/seaweedfs/seaweedfs/weed/s3api/s3err"
|
|
)
|
|
|
|
// BucketConfig represents cached bucket configuration
|
|
type BucketConfig struct {
|
|
Name string
|
|
Versioning string // "Enabled", "Suspended", or ""
|
|
Ownership string
|
|
ACL []byte
|
|
Owner string
|
|
IsPublicRead bool // Cached flag to avoid JSON parsing on every request
|
|
CORS *cors.CORSConfiguration
|
|
ObjectLockConfig *ObjectLockConfiguration // Cached parsed Object Lock configuration
|
|
BucketPolicy *policy_engine.PolicyDocument // Cached bucket policy for performance
|
|
// LifecycleTTL answers "what volume TTL should this PutObject get?"
|
|
// using only fast-path-safe predicates (prefix + size; tags excluded
|
|
// because they're mutable post-PUT). nil = no TTL applies (no
|
|
// lifecycle config, versioned bucket, or only ineligible rules).
|
|
// The full canonical rule set lives inside the resolver; the
|
|
// lifecycle worker reads bucket entries directly off the meta-log
|
|
// rather than this cache.
|
|
LifecycleTTL *LifecycleTTLResolver
|
|
KMSKeyCache *BucketKMSCache // Per-bucket KMS key cache for SSE-KMS operations
|
|
LastModified time.Time
|
|
Entry *filer_pb.Entry
|
|
}
|
|
|
|
// BucketKMSCache represents per-bucket KMS key caching for SSE-KMS operations
|
|
// This provides better isolation and automatic cleanup compared to global caching
|
|
type BucketKMSCache struct {
|
|
cache map[string]*BucketKMSCacheEntry // Key: contextHash, Value: cached data key
|
|
mutex sync.RWMutex
|
|
bucket string // Bucket name for logging/debugging
|
|
lastTTL time.Duration // TTL used for cache entries (typically 1 hour)
|
|
}
|
|
|
|
// BucketKMSCacheEntry represents a single cached KMS data key
|
|
type BucketKMSCacheEntry struct {
|
|
DataKey interface{} // Could be *kms.GenerateDataKeyResponse or similar
|
|
ExpiresAt time.Time
|
|
KeyID string
|
|
ContextHash string // Hash of encryption context for cache validation
|
|
}
|
|
|
|
// NewBucketKMSCache creates a new per-bucket KMS key cache
|
|
func NewBucketKMSCache(bucketName string, ttl time.Duration) *BucketKMSCache {
|
|
return &BucketKMSCache{
|
|
cache: make(map[string]*BucketKMSCacheEntry),
|
|
bucket: bucketName,
|
|
lastTTL: ttl,
|
|
}
|
|
}
|
|
|
|
// Get retrieves a cached KMS data key if it exists and hasn't expired
|
|
func (bkc *BucketKMSCache) Get(contextHash string) (*BucketKMSCacheEntry, bool) {
|
|
if bkc == nil {
|
|
return nil, false
|
|
}
|
|
|
|
bkc.mutex.RLock()
|
|
defer bkc.mutex.RUnlock()
|
|
|
|
entry, exists := bkc.cache[contextHash]
|
|
if !exists {
|
|
return nil, false
|
|
}
|
|
|
|
// Check if entry has expired
|
|
if time.Now().After(entry.ExpiresAt) {
|
|
return nil, false
|
|
}
|
|
|
|
return entry, true
|
|
}
|
|
|
|
// Set stores a KMS data key in the cache
|
|
func (bkc *BucketKMSCache) Set(contextHash, keyID string, dataKey interface{}, ttl time.Duration) {
|
|
if bkc == nil {
|
|
return
|
|
}
|
|
|
|
bkc.mutex.Lock()
|
|
defer bkc.mutex.Unlock()
|
|
|
|
bkc.cache[contextHash] = &BucketKMSCacheEntry{
|
|
DataKey: dataKey,
|
|
ExpiresAt: time.Now().Add(ttl),
|
|
KeyID: keyID,
|
|
ContextHash: contextHash,
|
|
}
|
|
bkc.lastTTL = ttl
|
|
}
|
|
|
|
// CleanupExpired removes expired entries from the cache
|
|
func (bkc *BucketKMSCache) CleanupExpired() int {
|
|
if bkc == nil {
|
|
return 0
|
|
}
|
|
|
|
bkc.mutex.Lock()
|
|
defer bkc.mutex.Unlock()
|
|
|
|
now := time.Now()
|
|
expiredCount := 0
|
|
|
|
for key, entry := range bkc.cache {
|
|
if now.After(entry.ExpiresAt) {
|
|
// Clear sensitive data before removing from cache
|
|
bkc.clearSensitiveData(entry)
|
|
delete(bkc.cache, key)
|
|
expiredCount++
|
|
}
|
|
}
|
|
|
|
return expiredCount
|
|
}
|
|
|
|
// Size returns the current number of cached entries
|
|
func (bkc *BucketKMSCache) Size() int {
|
|
if bkc == nil {
|
|
return 0
|
|
}
|
|
|
|
bkc.mutex.RLock()
|
|
defer bkc.mutex.RUnlock()
|
|
|
|
return len(bkc.cache)
|
|
}
|
|
|
|
// clearSensitiveData securely clears sensitive data from a cache entry
|
|
func (bkc *BucketKMSCache) clearSensitiveData(entry *BucketKMSCacheEntry) {
|
|
if dataKeyResp, ok := entry.DataKey.(*kms.GenerateDataKeyResponse); ok {
|
|
// Zero out the plaintext data key to prevent it from lingering in memory
|
|
if dataKeyResp.Plaintext != nil {
|
|
for i := range dataKeyResp.Plaintext {
|
|
dataKeyResp.Plaintext[i] = 0
|
|
}
|
|
dataKeyResp.Plaintext = nil
|
|
}
|
|
}
|
|
}
|
|
|
|
// Clear clears all cached KMS entries, securely zeroing sensitive data first
|
|
func (bkc *BucketKMSCache) Clear() {
|
|
if bkc == nil {
|
|
return
|
|
}
|
|
|
|
bkc.mutex.Lock()
|
|
defer bkc.mutex.Unlock()
|
|
|
|
// Clear sensitive data from all entries before deletion
|
|
for _, entry := range bkc.cache {
|
|
bkc.clearSensitiveData(entry)
|
|
}
|
|
|
|
// Clear the cache map
|
|
bkc.cache = make(map[string]*BucketKMSCacheEntry)
|
|
}
|
|
|
|
// BucketConfigCache provides caching for bucket configurations
|
|
// Cache entries are automatically updated/invalidated through metadata subscription events,
|
|
// so TTL serves as a safety fallback rather than the primary consistency mechanism
|
|
type BucketConfigCache struct {
|
|
cache map[string]*BucketConfig
|
|
negativeCache map[string]time.Time // Cache for non-existent buckets
|
|
mutex sync.RWMutex
|
|
ttl time.Duration // Safety fallback TTL; real-time consistency maintained via events
|
|
negativeTTL time.Duration // TTL for negative cache entries
|
|
}
|
|
|
|
// BucketMetadata represents the complete metadata for a bucket
|
|
type BucketMetadata struct {
|
|
Tags map[string]string `json:"tags,omitempty"`
|
|
CORS *cors.CORSConfiguration `json:"cors,omitempty"`
|
|
Encryption *s3_pb.EncryptionConfiguration `json:"encryption,omitempty"`
|
|
// Future extensions can be added here:
|
|
// Versioning *s3_pb.VersioningConfiguration `json:"versioning,omitempty"`
|
|
// Lifecycle *s3_pb.LifecycleConfiguration `json:"lifecycle,omitempty"`
|
|
// Notification *s3_pb.NotificationConfiguration `json:"notification,omitempty"`
|
|
// Replication *s3_pb.ReplicationConfiguration `json:"replication,omitempty"`
|
|
// Analytics *s3_pb.AnalyticsConfiguration `json:"analytics,omitempty"`
|
|
// Logging *s3_pb.LoggingConfiguration `json:"logging,omitempty"`
|
|
// Website *s3_pb.WebsiteConfiguration `json:"website,omitempty"`
|
|
// RequestPayer *s3_pb.RequestPayerConfiguration `json:"requestPayer,omitempty"`
|
|
// PublicAccess *s3_pb.PublicAccessConfiguration `json:"publicAccess,omitempty"`
|
|
}
|
|
|
|
// NewBucketMetadata creates a new BucketMetadata with default values
|
|
func NewBucketMetadata() *BucketMetadata {
|
|
return &BucketMetadata{
|
|
Tags: make(map[string]string),
|
|
}
|
|
}
|
|
|
|
// IsEmpty returns true if the metadata has no configuration set
|
|
func (bm *BucketMetadata) IsEmpty() bool {
|
|
return len(bm.Tags) == 0 && bm.CORS == nil && bm.Encryption == nil
|
|
}
|
|
|
|
// HasEncryption returns true if bucket has encryption configuration
|
|
func (bm *BucketMetadata) HasEncryption() bool {
|
|
return bm.Encryption != nil
|
|
}
|
|
|
|
// HasCORS returns true if bucket has CORS configuration
|
|
func (bm *BucketMetadata) HasCORS() bool {
|
|
return bm.CORS != nil
|
|
}
|
|
|
|
// HasTags returns true if bucket has tags
|
|
func (bm *BucketMetadata) HasTags() bool {
|
|
return len(bm.Tags) > 0
|
|
}
|
|
|
|
// NewBucketConfigCache creates a new bucket configuration cache
|
|
// TTL can be set to a longer duration since cache consistency is maintained
|
|
// through real-time metadata subscription events rather than TTL expiration
|
|
func NewBucketConfigCache(ttl time.Duration) *BucketConfigCache {
|
|
negativeTTL := ttl / 4 // Negative cache TTL is shorter than positive cache
|
|
if negativeTTL < 30*time.Second {
|
|
negativeTTL = 30 * time.Second // Minimum 30 seconds for negative cache
|
|
}
|
|
|
|
return &BucketConfigCache{
|
|
cache: make(map[string]*BucketConfig),
|
|
negativeCache: make(map[string]time.Time),
|
|
ttl: ttl,
|
|
negativeTTL: negativeTTL,
|
|
}
|
|
}
|
|
|
|
// Get retrieves bucket configuration from cache
|
|
func (bcc *BucketConfigCache) Get(bucket string) (*BucketConfig, bool) {
|
|
bcc.mutex.RLock()
|
|
defer bcc.mutex.RUnlock()
|
|
|
|
config, exists := bcc.cache[bucket]
|
|
if !exists {
|
|
return nil, false
|
|
}
|
|
|
|
// Check if cache entry is expired (safety fallback; entries are normally updated via events)
|
|
if time.Since(config.LastModified) > bcc.ttl {
|
|
return nil, false
|
|
}
|
|
|
|
return config, true
|
|
}
|
|
|
|
// Set stores bucket configuration in cache
|
|
func (bcc *BucketConfigCache) Set(bucket string, config *BucketConfig) {
|
|
bcc.mutex.Lock()
|
|
defer bcc.mutex.Unlock()
|
|
|
|
config.LastModified = time.Now()
|
|
bcc.cache[bucket] = config
|
|
}
|
|
|
|
// Remove removes bucket configuration from cache
|
|
func (bcc *BucketConfigCache) Remove(bucket string) {
|
|
bcc.mutex.Lock()
|
|
defer bcc.mutex.Unlock()
|
|
|
|
delete(bcc.cache, bucket)
|
|
}
|
|
|
|
// Clear clears all cached configurations
|
|
func (bcc *BucketConfigCache) Clear() {
|
|
bcc.mutex.Lock()
|
|
defer bcc.mutex.Unlock()
|
|
|
|
bcc.cache = make(map[string]*BucketConfig)
|
|
bcc.negativeCache = make(map[string]time.Time)
|
|
}
|
|
|
|
// IsNegativelyCached checks if a bucket is in the negative cache (doesn't exist)
|
|
func (bcc *BucketConfigCache) IsNegativelyCached(bucket string) bool {
|
|
bcc.mutex.Lock()
|
|
defer bcc.mutex.Unlock()
|
|
|
|
if cachedTime, exists := bcc.negativeCache[bucket]; exists {
|
|
// Check if the negative cache entry is still valid
|
|
if time.Since(cachedTime) < bcc.negativeTTL {
|
|
return true
|
|
}
|
|
// Entry expired, remove it
|
|
delete(bcc.negativeCache, bucket)
|
|
}
|
|
return false
|
|
}
|
|
|
|
// SetNegativeCache marks a bucket as non-existent in the negative cache
|
|
func (bcc *BucketConfigCache) SetNegativeCache(bucket string) {
|
|
bcc.mutex.Lock()
|
|
defer bcc.mutex.Unlock()
|
|
|
|
bcc.negativeCache[bucket] = time.Now()
|
|
}
|
|
|
|
// RemoveNegativeCache removes a bucket from the negative cache
|
|
func (bcc *BucketConfigCache) RemoveNegativeCache(bucket string) {
|
|
bcc.mutex.Lock()
|
|
defer bcc.mutex.Unlock()
|
|
|
|
delete(bcc.negativeCache, bucket)
|
|
}
|
|
|
|
// loadBucketPolicyFromExtended loads and parses bucket policy from entry extended attributes
|
|
func loadBucketPolicyFromExtended(entry *filer_pb.Entry, bucket string) *policy_engine.PolicyDocument {
|
|
if entry.Extended == nil {
|
|
return nil
|
|
}
|
|
|
|
policyJSON, exists := entry.Extended[BUCKET_POLICY_METADATA_KEY]
|
|
if !exists || len(policyJSON) == 0 {
|
|
glog.V(4).Infof("loadBucketPolicyFromExtended: no bucket policy found for bucket %s", bucket)
|
|
return nil
|
|
}
|
|
|
|
var policyDoc policy_engine.PolicyDocument
|
|
if err := json.Unmarshal(policyJSON, &policyDoc); err != nil {
|
|
glog.Errorf("loadBucketPolicyFromExtended: failed to parse bucket policy for %s: %v", bucket, err)
|
|
return nil
|
|
}
|
|
|
|
glog.V(3).Infof("loadBucketPolicyFromExtended: loaded bucket policy for bucket %s", bucket)
|
|
return &policyDoc
|
|
}
|
|
|
|
// getBucketConfig retrieves bucket configuration with caching
|
|
func (s3a *S3ApiServer) getBucketConfig(bucket string) (*BucketConfig, s3err.ErrorCode) {
|
|
// Check negative cache first
|
|
if s3a.bucketConfigCache.IsNegativelyCached(bucket) {
|
|
return nil, s3err.ErrNoSuchBucket
|
|
}
|
|
|
|
// Try positive cache
|
|
if config, found := s3a.bucketConfigCache.Get(bucket); found {
|
|
return config, s3err.ErrNone
|
|
}
|
|
|
|
// Try to get from filer
|
|
entry, err := s3a.getBucketEntry(bucket)
|
|
if err != nil {
|
|
if errors.Is(err, filer_pb.ErrNotFound) {
|
|
// Bucket doesn't exist - set negative cache
|
|
s3a.bucketConfigCache.SetNegativeCache(bucket)
|
|
return nil, s3err.ErrNoSuchBucket
|
|
}
|
|
glog.Errorf("getBucketConfig: failed to get bucket entry for %s: %v", bucket, err)
|
|
return nil, s3err.ErrInternalError
|
|
}
|
|
|
|
config := &BucketConfig{
|
|
Name: bucket,
|
|
Entry: entry,
|
|
IsPublicRead: false, // Explicitly default to false for private buckets
|
|
}
|
|
s3a.populateBucketConfigDerivedFields(config)
|
|
|
|
// Cache the result
|
|
s3a.bucketConfigCache.Set(bucket, config)
|
|
|
|
return config, s3err.ErrNone
|
|
}
|
|
|
|
// populateBucketConfigDerivedFields fills every field on BucketConfig that is
|
|
// derived from Entry.Extended / Entry.Content (versioning flag, ACL, owner,
|
|
// object lock, bucket policy, CORS, lifecycle TTL resolver). It is the
|
|
// single source of truth for that mapping; callers that take a fresh
|
|
// BucketConfig (getBucketConfig, updateBucketConfig after the user's update
|
|
// fn runs, the meta-log subscription cache refresher) all funnel through
|
|
// here so a missed field can't silently keep stale data — e.g. a stale
|
|
// LifecycleTTL after a Put/DeleteBucketLifecycle would keep stamping the
|
|
// old policy's irreversible volume TTL onto new writes.
|
|
func (s3a *S3ApiServer) populateBucketConfigDerivedFields(config *BucketConfig) {
|
|
// Reset every derived field so stale values from a previous Entry
|
|
// don't survive a clear (e.g. DELETE policy → BucketPolicy=nil).
|
|
config.Versioning = ""
|
|
config.Ownership = ""
|
|
config.ACL = nil
|
|
config.Owner = ""
|
|
config.IsPublicRead = false
|
|
config.ObjectLockConfig = nil
|
|
config.BucketPolicy = nil
|
|
config.LifecycleTTL = nil
|
|
config.CORS = nil
|
|
|
|
entry := config.Entry
|
|
if entry == nil {
|
|
return
|
|
}
|
|
bucket := config.Name
|
|
|
|
if entry.Extended != nil {
|
|
if versioning, exists := entry.Extended[s3_constants.ExtVersioningKey]; exists {
|
|
config.Versioning = string(versioning)
|
|
}
|
|
if ownership, exists := entry.Extended[s3_constants.ExtOwnershipKey]; exists {
|
|
config.Ownership = string(ownership)
|
|
}
|
|
if acl, exists := entry.Extended[s3_constants.ExtAmzAclKey]; exists {
|
|
config.ACL = acl
|
|
// Parse ACL once and cache public-read status.
|
|
config.IsPublicRead = parseAndCachePublicReadStatus(acl)
|
|
}
|
|
if owner, exists := entry.Extended[s3_constants.ExtAmzOwnerKey]; exists {
|
|
config.Owner = string(owner)
|
|
}
|
|
if objectLockConfig, found := LoadObjectLockConfigurationFromExtended(entry); found {
|
|
config.ObjectLockConfig = objectLockConfig
|
|
}
|
|
config.BucketPolicy = loadBucketPolicyFromExtended(entry, bucket)
|
|
|
|
// Pre-parse lifecycle XML so the per-write TTL resolver doesn't
|
|
// pay parsing cost on every PutObject. nil on parse error so
|
|
// the PUT path falls through to "no TTL" rather than rejecting
|
|
// writes.
|
|
if xmlBytes, ok := entry.Extended[bucketLifecycleConfigurationXMLKey]; ok && len(xmlBytes) > 0 {
|
|
if rules, err := lifecycle_xml.ParseCanonical(xmlBytes); err == nil {
|
|
// Object Lock requires versioning, so an ObjectLockConfig
|
|
// implies the bucket is versioned even when the explicit
|
|
// Versioning header was never written. BucketIsVersioned
|
|
// in this file uses the same OR — keep them aligned.
|
|
versioned := config.Versioning == s3_constants.VersioningEnabled ||
|
|
config.Versioning == s3_constants.VersioningSuspended ||
|
|
config.ObjectLockConfig != nil
|
|
config.LifecycleTTL = NewLifecycleTTLResolver(rules, versioned)
|
|
} else {
|
|
glog.V(1).Infof("populateBucketConfigDerivedFields: bucket %s lifecycle xml parse: %v", bucket, err)
|
|
}
|
|
}
|
|
}
|
|
|
|
// Sync bucket policy to the policy engine for evaluation.
|
|
s3a.syncBucketPolicyToEngine(bucket, config.BucketPolicy)
|
|
|
|
// Parse CORS configuration directly from the entry's Content field.
|
|
// This avoids a redundant RPC call since we already have the entry.
|
|
config.CORS = parseCORSFromEntryContent(entry.Content)
|
|
}
|
|
|
|
// updateBucketConfig updates bucket configuration and invalidates cache
|
|
func (s3a *S3ApiServer) updateBucketConfig(bucket string, updateFn func(*BucketConfig) error) s3err.ErrorCode {
|
|
config, errCode := s3a.getBucketConfig(bucket)
|
|
if errCode != s3err.ErrNone {
|
|
return errCode
|
|
}
|
|
|
|
nextConfig := cloneBucketConfig(config)
|
|
if nextConfig == nil {
|
|
glog.Errorf("updateBucketConfig: failed to clone config for bucket %s", bucket)
|
|
return s3err.ErrInternalError
|
|
}
|
|
|
|
// Apply update function
|
|
if err := updateFn(nextConfig); err != nil {
|
|
glog.Errorf("updateBucketConfig: update function failed for bucket %s: %v", bucket, err)
|
|
return s3err.ErrInternalError
|
|
}
|
|
|
|
// Prepare extended attributes
|
|
if nextConfig.Entry == nil {
|
|
glog.Errorf("updateBucketConfig: missing bucket entry for %s", bucket)
|
|
return s3err.ErrInternalError
|
|
}
|
|
if nextConfig.Entry.Extended == nil {
|
|
nextConfig.Entry.Extended = make(map[string][]byte)
|
|
}
|
|
|
|
// Update extended attributes
|
|
if nextConfig.Versioning != "" {
|
|
nextConfig.Entry.Extended[s3_constants.ExtVersioningKey] = []byte(nextConfig.Versioning)
|
|
} else {
|
|
delete(nextConfig.Entry.Extended, s3_constants.ExtVersioningKey)
|
|
}
|
|
if nextConfig.Ownership != "" {
|
|
nextConfig.Entry.Extended[s3_constants.ExtOwnershipKey] = []byte(nextConfig.Ownership)
|
|
} else {
|
|
delete(nextConfig.Entry.Extended, s3_constants.ExtOwnershipKey)
|
|
}
|
|
if nextConfig.ACL != nil {
|
|
nextConfig.Entry.Extended[s3_constants.ExtAmzAclKey] = nextConfig.ACL
|
|
} else {
|
|
delete(nextConfig.Entry.Extended, s3_constants.ExtAmzAclKey)
|
|
}
|
|
if nextConfig.Owner != "" {
|
|
nextConfig.Entry.Extended[s3_constants.ExtAmzOwnerKey] = []byte(nextConfig.Owner)
|
|
} else {
|
|
delete(nextConfig.Entry.Extended, s3_constants.ExtAmzOwnerKey)
|
|
}
|
|
// Update Object Lock configuration
|
|
if nextConfig.ObjectLockConfig != nil {
|
|
glog.V(3).Infof("updateBucketConfig: storing Object Lock config for bucket %s: %+v", bucket, nextConfig.ObjectLockConfig)
|
|
if err := StoreObjectLockConfigurationInExtended(nextConfig.Entry, nextConfig.ObjectLockConfig); err != nil {
|
|
glog.Errorf("updateBucketConfig: failed to store Object Lock configuration for bucket %s: %v", bucket, err)
|
|
return s3err.ErrInternalError
|
|
}
|
|
glog.V(3).Infof("updateBucketConfig: stored Object Lock config in extended attributes for bucket %s, key=%s, value=%s",
|
|
bucket, s3_constants.ExtObjectLockEnabledKey, string(nextConfig.Entry.Extended[s3_constants.ExtObjectLockEnabledKey]))
|
|
}
|
|
|
|
// Save to filer
|
|
glog.V(3).Infof("updateBucketConfig: saving entry to filer for bucket %s", bucket)
|
|
err := s3a.updateEntry(s3a.bucketRoot(bucket), nextConfig.Entry)
|
|
if err != nil {
|
|
glog.Errorf("updateBucketConfig: failed to update bucket entry for %s: %v", bucket, err)
|
|
return s3err.ErrInternalError
|
|
}
|
|
glog.V(3).Infof("updateBucketConfig: saved entry to filer for bucket %s", bucket)
|
|
|
|
// Update cache. Re-derive every Extended-backed field from the
|
|
// just-saved Entry — the user's update fn may have flipped, added,
|
|
// or cleared bytes (e.g. PutBucketLifecycle / DeleteBucketLifecycle
|
|
// rewrites the lifecycle XML key) and the resolver / parsed configs
|
|
// must follow.
|
|
s3a.populateBucketConfigDerivedFields(nextConfig)
|
|
s3a.bucketConfigCache.Set(bucket, nextConfig)
|
|
|
|
return s3err.ErrNone
|
|
}
|
|
|
|
func cloneBucketConfig(config *BucketConfig) *BucketConfig {
|
|
if config == nil {
|
|
return nil
|
|
}
|
|
|
|
cloned := *config
|
|
if config.ACL != nil {
|
|
cloned.ACL = append([]byte(nil), config.ACL...)
|
|
}
|
|
if config.Entry != nil {
|
|
cloned.Entry = proto.Clone(config.Entry).(*filer_pb.Entry)
|
|
}
|
|
if config.CORS != nil {
|
|
cloned.CORS = cloneCORSConfiguration(config.CORS)
|
|
}
|
|
if config.ObjectLockConfig != nil {
|
|
cloned.ObjectLockConfig = cloneObjectLockConfiguration(config.ObjectLockConfig)
|
|
}
|
|
if config.BucketPolicy != nil {
|
|
cloned.BucketPolicy = cloneBucketPolicy(config.BucketPolicy)
|
|
}
|
|
|
|
return &cloned
|
|
}
|
|
|
|
func cloneCORSConfiguration(config *cors.CORSConfiguration) *cors.CORSConfiguration {
|
|
if config == nil {
|
|
return nil
|
|
}
|
|
|
|
cloned := &cors.CORSConfiguration{
|
|
CORSRules: make([]cors.CORSRule, len(config.CORSRules)),
|
|
}
|
|
for i, rule := range config.CORSRules {
|
|
cloned.CORSRules[i] = cors.CORSRule{
|
|
AllowedHeaders: append([]string(nil), rule.AllowedHeaders...),
|
|
AllowedMethods: append([]string(nil), rule.AllowedMethods...),
|
|
AllowedOrigins: append([]string(nil), rule.AllowedOrigins...),
|
|
ExposeHeaders: append([]string(nil), rule.ExposeHeaders...),
|
|
ID: rule.ID,
|
|
}
|
|
if rule.MaxAgeSeconds != nil {
|
|
maxAge := *rule.MaxAgeSeconds
|
|
cloned.CORSRules[i].MaxAgeSeconds = &maxAge
|
|
}
|
|
}
|
|
|
|
return cloned
|
|
}
|
|
|
|
func cloneObjectLockConfiguration(config *ObjectLockConfiguration) *ObjectLockConfiguration {
|
|
if config == nil {
|
|
return nil
|
|
}
|
|
|
|
cloned := &ObjectLockConfiguration{
|
|
XMLNS: config.XMLNS,
|
|
XMLName: config.XMLName,
|
|
ObjectLockEnabled: config.ObjectLockEnabled,
|
|
}
|
|
if config.Rule != nil {
|
|
cloned.Rule = &ObjectLockRule{
|
|
XMLName: config.Rule.XMLName,
|
|
}
|
|
if config.Rule.DefaultRetention != nil {
|
|
cloned.Rule.DefaultRetention = &DefaultRetention{
|
|
XMLName: config.Rule.DefaultRetention.XMLName,
|
|
Mode: config.Rule.DefaultRetention.Mode,
|
|
Days: config.Rule.DefaultRetention.Days,
|
|
Years: config.Rule.DefaultRetention.Years,
|
|
DaysSet: config.Rule.DefaultRetention.DaysSet,
|
|
YearsSet: config.Rule.DefaultRetention.YearsSet,
|
|
}
|
|
}
|
|
}
|
|
|
|
return cloned
|
|
}
|
|
|
|
func cloneBucketPolicy(policyDoc *policy_engine.PolicyDocument) *policy_engine.PolicyDocument {
|
|
if policyDoc == nil {
|
|
return nil
|
|
}
|
|
|
|
cloned := &policy_engine.PolicyDocument{
|
|
Version: policyDoc.Version,
|
|
Statement: make([]policy_engine.PolicyStatement, len(policyDoc.Statement)),
|
|
}
|
|
for i, statement := range policyDoc.Statement {
|
|
cloned.Statement[i] = clonePolicyStatement(statement)
|
|
}
|
|
|
|
return cloned
|
|
}
|
|
|
|
func clonePolicyStatement(statement policy_engine.PolicyStatement) policy_engine.PolicyStatement {
|
|
cloned := policy_engine.PolicyStatement{
|
|
Sid: statement.Sid,
|
|
Effect: statement.Effect,
|
|
Action: cloneStringOrStringSlice(statement.Action),
|
|
NotResource: cloneStringOrStringSlicePtr(statement.NotResource),
|
|
Principal: cloneStringOrStringSlicePtr(statement.Principal),
|
|
Resource: cloneStringOrStringSlicePtr(statement.Resource),
|
|
}
|
|
if statement.Condition != nil {
|
|
cloned.Condition = make(policy_engine.PolicyConditions, len(statement.Condition))
|
|
for operator, operands := range statement.Condition {
|
|
copiedOperands := make(map[string]policy_engine.StringOrStringSlice, len(operands))
|
|
for key, value := range operands {
|
|
copiedOperands[key] = cloneStringOrStringSlice(value)
|
|
}
|
|
cloned.Condition[operator] = copiedOperands
|
|
}
|
|
}
|
|
return cloned
|
|
}
|
|
|
|
func cloneStringOrStringSlice(value policy_engine.StringOrStringSlice) policy_engine.StringOrStringSlice {
|
|
return policy_engine.CloneStringOrStringSlice(value)
|
|
}
|
|
|
|
func cloneStringOrStringSlicePtr(value *policy_engine.StringOrStringSlice) *policy_engine.StringOrStringSlice {
|
|
if value == nil {
|
|
return nil
|
|
}
|
|
cloned := policy_engine.CloneStringOrStringSlice(*value)
|
|
return &cloned
|
|
}
|
|
|
|
// isVersioningEnabled checks if versioning is enabled for a bucket (with caching)
|
|
func (s3a *S3ApiServer) isVersioningEnabled(bucket string) (bool, error) {
|
|
config, errCode := s3a.getBucketConfig(bucket)
|
|
if errCode != s3err.ErrNone {
|
|
if errCode == s3err.ErrNoSuchBucket {
|
|
return false, filer_pb.ErrNotFound
|
|
}
|
|
return false, fmt.Errorf("failed to get bucket config: %v", errCode)
|
|
}
|
|
|
|
// Versioning is enabled if explicitly set to "Enabled" OR if object lock is enabled
|
|
// (since object lock requires versioning to be enabled)
|
|
return config.Versioning == s3_constants.VersioningEnabled || config.ObjectLockConfig != nil, nil
|
|
}
|
|
|
|
// isVersioningConfigured checks if versioning has been configured (either Enabled or Suspended)
|
|
func (s3a *S3ApiServer) isVersioningConfigured(bucket string) (bool, error) {
|
|
config, errCode := s3a.getBucketConfig(bucket)
|
|
if errCode != s3err.ErrNone {
|
|
if errCode == s3err.ErrNoSuchBucket {
|
|
return false, filer_pb.ErrNotFound
|
|
}
|
|
return false, fmt.Errorf("failed to get bucket config: %v", errCode)
|
|
}
|
|
|
|
// Versioning is configured if explicitly set to either "Enabled" or "Suspended"
|
|
// OR if object lock is enabled (which forces versioning)
|
|
return config.Versioning != "" || config.ObjectLockConfig != nil, nil
|
|
}
|
|
|
|
// isObjectLockEnabled checks if Object Lock is enabled for a bucket (with caching)
|
|
func (s3a *S3ApiServer) isObjectLockEnabled(bucket string) (bool, error) {
|
|
config, errCode := s3a.getBucketConfig(bucket)
|
|
if errCode != s3err.ErrNone {
|
|
if errCode == s3err.ErrNoSuchBucket {
|
|
return false, filer_pb.ErrNotFound
|
|
}
|
|
return false, fmt.Errorf("failed to get bucket config: %v", errCode)
|
|
}
|
|
|
|
return config.ObjectLockConfig != nil, nil
|
|
}
|
|
|
|
// getVersioningState returns the detailed versioning state for a bucket
|
|
func (s3a *S3ApiServer) getVersioningState(bucket string) (string, error) {
|
|
config, errCode := s3a.getBucketConfig(bucket)
|
|
if errCode != s3err.ErrNone {
|
|
if errCode == s3err.ErrNoSuchBucket {
|
|
// Signal to callers that the bucket does not exist so they can
|
|
// decide whether to auto-create it (e.g., in PUT handlers).
|
|
return "", filer_pb.ErrNotFound
|
|
}
|
|
glog.Errorf("getVersioningState: failed to get bucket config for %s: %v", bucket, errCode)
|
|
return "", fmt.Errorf("failed to get bucket config: %v", errCode)
|
|
}
|
|
|
|
// If object lock is enabled, versioning must be enabled regardless of explicit setting
|
|
if config.ObjectLockConfig != nil {
|
|
return s3_constants.VersioningEnabled, nil
|
|
}
|
|
|
|
// Return the explicit versioning status (empty string means never configured)
|
|
return config.Versioning, nil
|
|
}
|
|
|
|
// getBucketVersioningStatus returns the versioning status for a bucket
|
|
func (s3a *S3ApiServer) getBucketVersioningStatus(bucket string) (string, s3err.ErrorCode) {
|
|
config, errCode := s3a.getBucketConfig(bucket)
|
|
if errCode != s3err.ErrNone {
|
|
return "", errCode
|
|
}
|
|
|
|
// Return exactly what's stored - empty string means versioning was never configured
|
|
// This matches AWS S3 behavior where new buckets have no Status field in GetBucketVersioning response
|
|
return config.Versioning, s3err.ErrNone
|
|
}
|
|
|
|
// setBucketVersioningStatus sets the versioning status for a bucket
|
|
func (s3a *S3ApiServer) setBucketVersioningStatus(bucket, status string) s3err.ErrorCode {
|
|
errCode := s3a.updateBucketConfig(bucket, func(config *BucketConfig) error {
|
|
config.Versioning = status
|
|
return nil
|
|
})
|
|
return errCode
|
|
}
|
|
|
|
// getBucketOwnership returns the ownership setting for a bucket
|
|
func (s3a *S3ApiServer) getBucketOwnership(bucket string) (string, s3err.ErrorCode) {
|
|
config, errCode := s3a.getBucketConfig(bucket)
|
|
if errCode != s3err.ErrNone {
|
|
return "", errCode
|
|
}
|
|
|
|
return config.Ownership, s3err.ErrNone
|
|
}
|
|
|
|
// setBucketOwnership sets the ownership setting for a bucket
|
|
func (s3a *S3ApiServer) setBucketOwnership(bucket, ownership string) s3err.ErrorCode {
|
|
return s3a.updateBucketConfig(bucket, func(config *BucketConfig) error {
|
|
config.Ownership = ownership
|
|
return nil
|
|
})
|
|
}
|
|
|
|
// parseCORSFromEntryContent parses CORS configuration directly from an entry's Content field.
|
|
// This avoids a separate RPC call when the entry is already available (e.g., from a
|
|
// subscription event or a prior getBucketEntry call).
|
|
func parseCORSFromEntryContent(content []byte) *cors.CORSConfiguration {
|
|
if len(content) == 0 {
|
|
return nil
|
|
}
|
|
var protoMetadata s3_pb.BucketMetadata
|
|
if err := proto.Unmarshal(content, &protoMetadata); err != nil {
|
|
glog.Errorf("parseCORSFromEntryContent: failed to unmarshal protobuf metadata: %v", err)
|
|
return nil
|
|
}
|
|
return corsConfigFromProto(protoMetadata.Cors)
|
|
}
|
|
|
|
// getCORSConfiguration retrieves CORS configuration with caching
|
|
func (s3a *S3ApiServer) getCORSConfiguration(bucket string) (*cors.CORSConfiguration, s3err.ErrorCode) {
|
|
config, errCode := s3a.getBucketConfig(bucket)
|
|
if errCode != s3err.ErrNone {
|
|
return nil, errCode
|
|
}
|
|
|
|
return config.CORS, s3err.ErrNone
|
|
}
|
|
|
|
// updateCORSConfiguration updates the CORS configuration for a bucket
|
|
func (s3a *S3ApiServer) updateCORSConfiguration(bucket string, corsConfig *cors.CORSConfiguration) s3err.ErrorCode {
|
|
// Update using structured API
|
|
// Note: UpdateBucketCORS -> UpdateBucketMetadata -> setBucketMetadata
|
|
// already invalidates the cache synchronously after successful update
|
|
err := s3a.UpdateBucketCORS(bucket, corsConfig)
|
|
if err != nil {
|
|
glog.Errorf("updateCORSConfiguration: failed to update CORS config for bucket %s: %v", bucket, err)
|
|
return s3err.ErrInternalError
|
|
}
|
|
|
|
return s3err.ErrNone
|
|
}
|
|
|
|
// removeCORSConfiguration removes the CORS configuration for a bucket
|
|
func (s3a *S3ApiServer) removeCORSConfiguration(bucket string) s3err.ErrorCode {
|
|
// Update using structured API
|
|
// Note: ClearBucketCORS -> UpdateBucketMetadata -> setBucketMetadata
|
|
// already invalidates the cache synchronously after successful update
|
|
err := s3a.ClearBucketCORS(bucket)
|
|
if err != nil {
|
|
glog.Errorf("removeCORSConfiguration: failed to remove CORS config for bucket %s: %v", bucket, err)
|
|
return s3err.ErrInternalError
|
|
}
|
|
|
|
return s3err.ErrNone
|
|
}
|
|
|
|
// Conversion functions between CORS types and protobuf types
|
|
|
|
// corsRuleToProto converts a CORS rule to protobuf format
|
|
func corsRuleToProto(rule cors.CORSRule) *s3_pb.CORSRule {
|
|
return &s3_pb.CORSRule{
|
|
AllowedHeaders: rule.AllowedHeaders,
|
|
AllowedMethods: rule.AllowedMethods,
|
|
AllowedOrigins: rule.AllowedOrigins,
|
|
ExposeHeaders: rule.ExposeHeaders,
|
|
MaxAgeSeconds: int32(getMaxAgeSecondsValue(rule.MaxAgeSeconds)),
|
|
Id: rule.ID,
|
|
}
|
|
}
|
|
|
|
// corsRuleFromProto converts a protobuf CORS rule to standard format
|
|
func corsRuleFromProto(protoRule *s3_pb.CORSRule) cors.CORSRule {
|
|
var maxAge *int
|
|
// Always create the pointer if MaxAgeSeconds is >= 0
|
|
// This prevents nil pointer dereferences in tests and matches AWS behavior
|
|
if protoRule.MaxAgeSeconds >= 0 {
|
|
age := int(protoRule.MaxAgeSeconds)
|
|
maxAge = &age
|
|
}
|
|
// Only leave maxAge as nil if MaxAgeSeconds was explicitly set to a negative value
|
|
|
|
return cors.CORSRule{
|
|
AllowedHeaders: protoRule.AllowedHeaders,
|
|
AllowedMethods: protoRule.AllowedMethods,
|
|
AllowedOrigins: protoRule.AllowedOrigins,
|
|
ExposeHeaders: protoRule.ExposeHeaders,
|
|
MaxAgeSeconds: maxAge,
|
|
ID: protoRule.Id,
|
|
}
|
|
}
|
|
|
|
// corsConfigToProto converts CORS configuration to protobuf format
|
|
func corsConfigToProto(config *cors.CORSConfiguration) *s3_pb.CORSConfiguration {
|
|
if config == nil {
|
|
return nil
|
|
}
|
|
|
|
protoRules := make([]*s3_pb.CORSRule, len(config.CORSRules))
|
|
for i, rule := range config.CORSRules {
|
|
protoRules[i] = corsRuleToProto(rule)
|
|
}
|
|
|
|
return &s3_pb.CORSConfiguration{
|
|
CorsRules: protoRules,
|
|
}
|
|
}
|
|
|
|
// corsConfigFromProto converts protobuf CORS configuration to standard format
|
|
func corsConfigFromProto(protoConfig *s3_pb.CORSConfiguration) *cors.CORSConfiguration {
|
|
if protoConfig == nil {
|
|
return nil
|
|
}
|
|
|
|
rules := make([]cors.CORSRule, len(protoConfig.CorsRules))
|
|
for i, protoRule := range protoConfig.CorsRules {
|
|
rules[i] = corsRuleFromProto(protoRule)
|
|
}
|
|
|
|
return &cors.CORSConfiguration{
|
|
CORSRules: rules,
|
|
}
|
|
}
|
|
|
|
// getMaxAgeSecondsValue safely extracts max age seconds value
|
|
func getMaxAgeSecondsValue(maxAge *int) int {
|
|
if maxAge == nil {
|
|
return 0
|
|
}
|
|
return *maxAge
|
|
}
|
|
|
|
// parseAndCachePublicReadStatus parses the ACL and caches the public-read status
|
|
func parseAndCachePublicReadStatus(acl []byte) bool {
|
|
var grants []*s3.Grant
|
|
if err := json.Unmarshal(acl, &grants); err != nil {
|
|
return false
|
|
}
|
|
|
|
// Check if any grant gives read permission to "AllUsers" group
|
|
for _, grant := range grants {
|
|
if grant.Grantee != nil && grant.Grantee.URI != nil && grant.Permission != nil {
|
|
// Check for AllUsers group with Read permission
|
|
if *grant.Grantee.URI == s3_constants.GranteeGroupAllUsers &&
|
|
(*grant.Permission == s3_constants.PermissionRead || *grant.Permission == s3_constants.PermissionFullControl) {
|
|
return true
|
|
}
|
|
}
|
|
}
|
|
|
|
return false
|
|
}
|
|
|
|
// getBucketMetadata retrieves bucket metadata as a structured object with caching
|
|
func (s3a *S3ApiServer) getBucketMetadata(bucket string) (*BucketMetadata, error) {
|
|
if s3a.bucketConfigCache != nil {
|
|
// Check negative cache first
|
|
if s3a.bucketConfigCache.IsNegativelyCached(bucket) {
|
|
return nil, fmt.Errorf("bucket directory not found %s", bucket)
|
|
}
|
|
|
|
// Try to get from positive cache
|
|
if config, found := s3a.bucketConfigCache.Get(bucket); found {
|
|
// Extract metadata from cached config
|
|
if metadata, err := s3a.extractMetadataFromConfig(config); err == nil {
|
|
return metadata, nil
|
|
}
|
|
// If extraction fails, fall through to direct load
|
|
}
|
|
}
|
|
|
|
// Load directly from filer
|
|
return s3a.loadBucketMetadataFromFiler(bucket)
|
|
}
|
|
|
|
// extractMetadataFromConfig extracts BucketMetadata from cached BucketConfig
|
|
func (s3a *S3ApiServer) extractMetadataFromConfig(config *BucketConfig) (*BucketMetadata, error) {
|
|
if config == nil || config.Entry == nil {
|
|
return NewBucketMetadata(), nil
|
|
}
|
|
|
|
// Parse metadata from entry content if available
|
|
if len(config.Entry.Content) > 0 {
|
|
var protoMetadata s3_pb.BucketMetadata
|
|
if err := proto.Unmarshal(config.Entry.Content, &protoMetadata); err != nil {
|
|
glog.Errorf("extractMetadataFromConfig: failed to unmarshal protobuf metadata for bucket %s: %v", config.Name, err)
|
|
return nil, err
|
|
}
|
|
// Convert protobuf to structured metadata
|
|
metadata := &BucketMetadata{
|
|
Tags: protoMetadata.Tags,
|
|
CORS: corsConfigFromProto(protoMetadata.Cors),
|
|
Encryption: protoMetadata.Encryption,
|
|
}
|
|
return metadata, nil
|
|
}
|
|
|
|
// Fallback: create metadata from cached CORS config
|
|
metadata := NewBucketMetadata()
|
|
if config.CORS != nil {
|
|
metadata.CORS = config.CORS
|
|
}
|
|
|
|
return metadata, nil
|
|
}
|
|
|
|
// loadBucketMetadataFromFiler loads bucket metadata directly from the filer
|
|
func (s3a *S3ApiServer) loadBucketMetadataFromFiler(bucket string) (*BucketMetadata, error) {
|
|
// Validate bucket name to prevent path traversal attacks
|
|
if bucket == "" || strings.Contains(bucket, "/") || strings.Contains(bucket, "\\") ||
|
|
strings.Contains(bucket, "..") || strings.Contains(bucket, "~") {
|
|
return nil, fmt.Errorf("invalid bucket name: %s", bucket)
|
|
}
|
|
|
|
// Clean the bucket name further to prevent any potential path traversal
|
|
bucket = filepath.Clean(bucket)
|
|
if bucket == "." || bucket == ".." {
|
|
return nil, fmt.Errorf("invalid bucket name: %s", bucket)
|
|
}
|
|
|
|
// Get bucket directory entry to access its content
|
|
entry, err := s3a.getBucketEntry(bucket)
|
|
if err != nil {
|
|
// Check if this is a "not found" error
|
|
if errors.Is(err, filer_pb.ErrNotFound) {
|
|
// Set negative cache for non-existent bucket
|
|
if s3a.bucketConfigCache != nil {
|
|
s3a.bucketConfigCache.SetNegativeCache(bucket)
|
|
}
|
|
}
|
|
return nil, fmt.Errorf("error retrieving bucket directory %s: %w", bucket, err)
|
|
}
|
|
if entry == nil {
|
|
// Set negative cache for non-existent bucket
|
|
if s3a.bucketConfigCache != nil {
|
|
s3a.bucketConfigCache.SetNegativeCache(bucket)
|
|
}
|
|
return nil, fmt.Errorf("bucket directory not found %s", bucket)
|
|
}
|
|
|
|
// If no content, return empty metadata
|
|
if len(entry.Content) == 0 {
|
|
return NewBucketMetadata(), nil
|
|
}
|
|
|
|
// Unmarshal metadata from protobuf
|
|
var protoMetadata s3_pb.BucketMetadata
|
|
if err := proto.Unmarshal(entry.Content, &protoMetadata); err != nil {
|
|
glog.Errorf("getBucketMetadata: failed to unmarshal protobuf metadata for bucket %s: %v", bucket, err)
|
|
return nil, fmt.Errorf("failed to unmarshal bucket metadata for %s: %w", bucket, err)
|
|
}
|
|
|
|
// Convert protobuf CORS to standard CORS
|
|
corsConfig := corsConfigFromProto(protoMetadata.Cors)
|
|
|
|
// Create and return structured metadata
|
|
metadata := &BucketMetadata{
|
|
Tags: protoMetadata.Tags,
|
|
CORS: corsConfig,
|
|
Encryption: protoMetadata.Encryption,
|
|
}
|
|
|
|
return metadata, nil
|
|
}
|
|
|
|
// setBucketMetadata stores bucket metadata from a structured object
|
|
func (s3a *S3ApiServer) setBucketMetadata(bucket string, metadata *BucketMetadata) error {
|
|
// Validate bucket name to prevent path traversal attacks
|
|
if bucket == "" || strings.Contains(bucket, "/") || strings.Contains(bucket, "\\") ||
|
|
strings.Contains(bucket, "..") || strings.Contains(bucket, "~") {
|
|
return fmt.Errorf("invalid bucket name: %s", bucket)
|
|
}
|
|
|
|
// Clean the bucket name further to prevent any potential path traversal
|
|
bucket = filepath.Clean(bucket)
|
|
if bucket == "." || bucket == ".." {
|
|
return fmt.Errorf("invalid bucket name: %s", bucket)
|
|
}
|
|
|
|
// Default to empty metadata if nil
|
|
if metadata == nil {
|
|
metadata = NewBucketMetadata()
|
|
}
|
|
|
|
// Create protobuf metadata
|
|
protoMetadata := &s3_pb.BucketMetadata{
|
|
Tags: metadata.Tags,
|
|
Cors: corsConfigToProto(metadata.CORS),
|
|
Encryption: metadata.Encryption,
|
|
}
|
|
|
|
// Marshal metadata to protobuf
|
|
metadataBytes, err := proto.Marshal(protoMetadata)
|
|
if err != nil {
|
|
return fmt.Errorf("failed to marshal bucket metadata to protobuf: %w", err)
|
|
}
|
|
|
|
// Update the bucket entry with new content
|
|
err = s3a.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
|
|
// Get current bucket entry
|
|
entry, err := s3a.getBucketEntry(bucket)
|
|
if err != nil {
|
|
return fmt.Errorf("error retrieving bucket directory %s: %w", bucket, err)
|
|
}
|
|
if entry == nil {
|
|
return fmt.Errorf("bucket directory not found %s", bucket)
|
|
}
|
|
|
|
// Update content with metadata
|
|
entry.Content = metadataBytes
|
|
|
|
request := &filer_pb.UpdateEntryRequest{
|
|
Directory: s3a.bucketRoot(bucket),
|
|
Entry: entry,
|
|
}
|
|
|
|
_, err = client.UpdateEntry(context.Background(), request)
|
|
return err
|
|
})
|
|
|
|
// Invalidate cache after successful update
|
|
if err == nil && s3a.bucketConfigCache != nil {
|
|
s3a.bucketConfigCache.Remove(bucket)
|
|
s3a.bucketConfigCache.RemoveNegativeCache(bucket) // Remove from negative cache too
|
|
}
|
|
|
|
return err
|
|
}
|
|
|
|
// New structured API functions using BucketMetadata
|
|
|
|
// GetBucketMetadata retrieves complete bucket metadata as a structured object
|
|
func (s3a *S3ApiServer) GetBucketMetadata(bucket string) (*BucketMetadata, error) {
|
|
return s3a.getBucketMetadata(bucket)
|
|
}
|
|
|
|
// SetBucketMetadata stores complete bucket metadata from a structured object
|
|
func (s3a *S3ApiServer) SetBucketMetadata(bucket string, metadata *BucketMetadata) error {
|
|
return s3a.setBucketMetadata(bucket, metadata)
|
|
}
|
|
|
|
// UpdateBucketMetadata updates specific parts of bucket metadata while preserving others
|
|
//
|
|
// DISTRIBUTED SYSTEM DESIGN NOTE:
|
|
// This function implements a read-modify-write pattern with "last write wins" semantics.
|
|
// In the rare case of concurrent updates to different parts of bucket metadata
|
|
// (e.g., simultaneous tag and CORS updates), the last write may overwrite previous changes.
|
|
//
|
|
// This is an acceptable trade-off because:
|
|
// 1. Bucket metadata updates are infrequent in typical S3 usage
|
|
// 2. Traditional locking doesn't work in distributed systems across multiple nodes
|
|
// 3. The complexity of distributed consensus (e.g., Raft) for metadata updates would
|
|
// be disproportionate to the low frequency of bucket configuration changes
|
|
// 4. Most bucket operations (tags, CORS, encryption) are typically configured once
|
|
// during setup rather than being frequently modified
|
|
//
|
|
// If stronger consistency is required, consider implementing optimistic concurrency
|
|
// control with version numbers or ETags at the storage layer.
|
|
func (s3a *S3ApiServer) UpdateBucketMetadata(bucket string, update func(*BucketMetadata) error) error {
|
|
// Get current metadata
|
|
metadata, err := s3a.GetBucketMetadata(bucket)
|
|
if err != nil {
|
|
return fmt.Errorf("failed to get current bucket metadata: %w", err)
|
|
}
|
|
|
|
// Apply update function
|
|
if err := update(metadata); err != nil {
|
|
return fmt.Errorf("failed to apply metadata update: %w", err)
|
|
}
|
|
|
|
// Store updated metadata (last write wins)
|
|
return s3a.SetBucketMetadata(bucket, metadata)
|
|
}
|
|
|
|
// Helper functions for specific metadata operations using structured API
|
|
|
|
// UpdateBucketTags sets bucket tags using the structured API
|
|
func (s3a *S3ApiServer) UpdateBucketTags(bucket string, tags map[string]string) error {
|
|
return s3a.UpdateBucketMetadata(bucket, func(metadata *BucketMetadata) error {
|
|
metadata.Tags = tags
|
|
return nil
|
|
})
|
|
}
|
|
|
|
// UpdateBucketCORS sets bucket CORS configuration using the structured API
|
|
func (s3a *S3ApiServer) UpdateBucketCORS(bucket string, corsConfig *cors.CORSConfiguration) error {
|
|
return s3a.UpdateBucketMetadata(bucket, func(metadata *BucketMetadata) error {
|
|
metadata.CORS = corsConfig
|
|
return nil
|
|
})
|
|
}
|
|
|
|
// UpdateBucketEncryption sets bucket encryption configuration using the structured API
|
|
func (s3a *S3ApiServer) UpdateBucketEncryption(bucket string, encryptionConfig *s3_pb.EncryptionConfiguration) error {
|
|
return s3a.UpdateBucketMetadata(bucket, func(metadata *BucketMetadata) error {
|
|
metadata.Encryption = encryptionConfig
|
|
return nil
|
|
})
|
|
}
|
|
|
|
// ClearBucketTags removes all bucket tags using the structured API
|
|
func (s3a *S3ApiServer) ClearBucketTags(bucket string) error {
|
|
return s3a.UpdateBucketMetadata(bucket, func(metadata *BucketMetadata) error {
|
|
metadata.Tags = make(map[string]string)
|
|
return nil
|
|
})
|
|
}
|
|
|
|
// ClearBucketCORS removes bucket CORS configuration using the structured API
|
|
func (s3a *S3ApiServer) ClearBucketCORS(bucket string) error {
|
|
return s3a.UpdateBucketMetadata(bucket, func(metadata *BucketMetadata) error {
|
|
metadata.CORS = nil
|
|
return nil
|
|
})
|
|
}
|
|
|
|
// ClearBucketEncryption removes bucket encryption configuration using the structured API
|
|
func (s3a *S3ApiServer) ClearBucketEncryption(bucket string) error {
|
|
return s3a.UpdateBucketMetadata(bucket, func(metadata *BucketMetadata) error {
|
|
metadata.Encryption = nil
|
|
return nil
|
|
})
|
|
}
|