seaweedfs

Commit Graph

Author	SHA1	Message	Date
Chris Lu	737116e83c	fix port probing	6 days ago
Chris Lu	07f3f5eec5	remove worker links	6 days ago
Chris Lu	47cad59c70	Remove misleading Workers sub-menu items from admin sidebar (#8594 ) * Remove misleading Workers sub-menu items from admin sidebar The sidebar sub-items (Job Detection, Job Queue, Job Execution, Configuration) always navigated to the first job type's tabs (typically EC Encoding) rather than showing cross-job-type views. This was confusing as noted in #8590. Since the in-page tabs already provide this navigation, remove the redundant sidebar sub-items and keep only the top-level Workers link. Fixes #8590 * Update layout_templ.go	6 days ago
Chris Lu	b17e2b411a	Add dynamic timeouts to plugin worker vacuum gRPC calls (#8593 ) * add dynamic timeouts to plugin worker vacuum gRPC calls All vacuum gRPC calls used context.Background() with no deadline, so the plugin scheduler's execution timeout could kill a job while a large volume compact was still in progress. Use volume-size-scaled timeouts matching the topology vacuum approach: 3 min/GB for compact, 1 min/GB for check, commit, and cleanup. Fixes #8591 * scale scheduler execution timeout by volume size The scheduler's per-job execution timeout (default 240s) would kill vacuum jobs on large volumes before they finish. Three changes: 1. Vacuum detection now includes estimated_runtime_seconds in job proposals, computed as 5 min/GB of volume size. 2. The scheduler checks for estimated_runtime_seconds in job parameters and uses it as the execution timeout when larger than the default — a generic mechanism any handler can use. 3. Vacuum task gRPC calls now use the passed-in ctx as parent instead of context.Background(), so scheduler cancellation propagates to in-flight RPCs. * extend job type runtime when proposals need more time The JobTypeMaxRuntime (default 30 min) wraps both detection and execution. Its context is the parent of all per-job execution contexts, so even with per-job estimated_runtime_seconds, jobCtx would cancel everything when it expires. After detection, scan proposals for the maximum estimated_runtime_seconds. If any proposal needs more time than the remaining JobTypeMaxRuntime, create a new execution context with enough headroom. This lets large vacuum jobs complete without being killed by the job type deadline while still respecting the configured limit for normal-sized jobs. * log missing volume size metric, remove dead minimum runtime guard Add a debug log in vacuumTimeout when t.volumeSize is 0 so operators can investigate why metrics are missing for a volume. Remove the unreachable estimatedRuntimeSeconds < 180 check in buildVacuumProposal — volumeSizeGB always >= 1 (due to +1 floor), so estimatedRuntimeSeconds is always >= 300. * cap estimated runtime and fix status check context - Cap maxEstimatedRuntime and per-job timeout overrides to 8 hours to prevent unbounded timeouts from bad metrics. - Check execCtx.Err() instead of jobCtx.Err() for status reporting, since dispatch runs under execCtx which may have a longer deadline. A successful dispatch under execCtx was misreported as "timeout" when jobCtx had expired.	6 days ago
Chris Lu	4c88fbfd5e	Fix nil pointer crash during concurrent vacuum compaction (#8592 ) * check for nil needle map before compaction sync When CommitCompact runs concurrently, it sets v.nm = nil under dataFileAccessLock. CompactByIndex does not hold that lock, so v.nm.Sync() can hit a nil pointer. Add an early nil check to return an error instead of crashing. Fixes #8591 * guard copyDataBasedOnIndexFile size check against nil needle map The post-compaction size validation at line 538 accesses v.nm.ContentSize() and v.nm.DeletedSize(). If CommitCompact has concurrently set v.nm to nil, this causes a SIGSEGV. Skip the validation when v.nm is nil since the actual data copy uses local needle maps (oldNm/newNm) and is unaffected. Fixes #8591 * use atomic.Bool for compaction flags to prevent concurrent vacuum races The isCompacting and isCommitCompacting flags were plain bools read and written from multiple goroutines without synchronization. This allowed concurrent vacuums on the same volume to pass the guard checks and run simultaneously, leading to the nil pointer crash. Using atomic.Bool with CompareAndSwap ensures only one compaction or commit can run per volume at a time. Fixes #8591 * use go-version-file in CI workflows instead of hardcoded versions Use go-version-file: 'go.mod' so CI automatically picks up the Go version from go.mod, avoiding future version drift. Reordered checkout before setup-go in go.yml and e2e.yml so go.mod is available. Removed the now-unused GO_VERSION env vars. * capture v.nm locally in CompactByIndex to close TOCTOU race A bare nil check on v.nm followed by v.nm.Sync() has a race window where CommitCompact can set v.nm = nil between the two. Snapshot the pointer into a local variable so the nil check and Sync operate on the same reference. * add dynamic timeouts to plugin worker vacuum gRPC calls All vacuum gRPC calls used context.Background() with no deadline, so the plugin scheduler's execution timeout could kill a job while a large volume compact was still in progress. Use volume-size-scaled timeouts matching the topology vacuum approach: 3 min/GB for compact, 1 min/GB for check, commit, and cleanup. Fixes #8591 * Revert "add dynamic timeouts to plugin worker vacuum gRPC calls" This reverts commit `80951934c3`. * unify compaction lifecycle into single atomic flag Replace separate isCompacting and isCommitCompacting flags with a single isCompactionInProgress atomic.Bool. This ensures CompactBy*, CommitCompact, Close, and Destroy are mutually exclusive — only one can run at a time per volume. Key changes: - All entry points use CompareAndSwap(false, true) to claim exclusive access. CompactByVolumeData and CompactByIndex now also guard v.nm and v.DataBackend with local captures. - Close() waits for the flag outside dataFileAccessLock to avoid deadlocking with CommitCompact (which holds the flag while waiting for the lock). It claims the flag before acquiring the lock so no new compaction can start. - Destroy() uses CAS instead of a racy Load check, preventing concurrent compaction from racing with volume teardown. - unmountVolumeByCollection no longer deletes from the map; DeleteCollectionFromDiskLocation removes entries only after successful Destroy, preventing orphaned volumes on failure. Fixes #8591	6 days ago
Chris Lu	d4d2e511ed	for mini, default to bind all	6 days ago
Chris Lu	d89a78d9e3	reduce logs	6 days ago
Chris Lu	00000ec006	Update s3_buckets_templ.go	6 days ago
Chris Lu	1bd7a98a4a	simplify plugin scheduler: remove configurable IdleSleepSeconds, use constant 61s The SchedulerConfig struct and its persistence/API were unnecessary indirection. Replace with a simple constant (reduced from 613s to 61s) so the scheduler re-checks for detectable job types promptly after going idle, improving the clean-install experience.	6 days ago
Chris Lu	8ad58e7002	4.16	6 days ago
Chris Lu	cf3693651c	fix: add IdxFileSize check to pre-delete volume verification The verification step checked DatFileSize and FileCount but not IdxFileSize, leaving a gap in the copy validation before source deletion.	7 days ago
Chris Lu	5f85bf5e8a	Batch volume balance: run multiple moves per job (#8561 ) * proto: add BalanceMoveSpec and batch fields to BalanceTaskParams Add BalanceMoveSpec message for encoding individual volume moves, and max_concurrent_moves + repeated moves fields to BalanceTaskParams to support batching multiple volume moves in a single job. * balance handler: add batch execution with concurrent volume moves Refactor Execute() into executeSingleMove() (backward compatible) and executeBatchMoves() which runs multiple volume moves concurrently using a semaphore-bounded goroutine pool. When BalanceTaskParams.Moves is populated, the batch path is taken; otherwise the single-move path. Includes aggregate progress reporting across concurrent moves, per-move error collection, and partial failure support. * balance handler: add batch config fields to Descriptor and worker config Add max_concurrent_moves and batch_size fields to the worker config form and deriveBalanceWorkerConfig(). These control how many volume moves run concurrently within a batch job and the maximum batch size. * balance handler: group detection proposals into batch jobs When batch_size > 1, the Detect method groups detection results into batch proposals where each proposal encodes multiple BalanceMoveSpec entries in BalanceTaskParams.Moves. Single-result batches fall back to the existing single-move proposal format for backward compatibility. * admin UI: add volume balance execution plan and batch badge Add renderBalanceExecutionPlan() for rich rendering of volume balance jobs in the job detail modal. Single-move jobs show source/target/volume info; batch jobs show a moves table with all volume moves. Add batch badge (e.g., "5 moves") next to job type in the execution jobs table when the job has batch=true label. * Update plugin_templ.go * fix: detection algorithm uses greedy target instead of divergent topology scores The detection loop tracked effective volume counts via an adjustments map, but createBalanceTask independently called planBalanceDestination which used the topology's LoadCount — a separate, unadjusted source of truth. This divergence caused multiple moves to pile onto the same server. Changes: - Add resolveBalanceDestination to resolve the detection loop's greedy target (minServer) rather than independently picking a destination - Add oscillation guard: stop when max-min <= 1 since no single move can improve the balance beyond that point - Track unseeded destinations: if a target server wasn't in the initial serverVolumeCounts, add it so subsequent iterations include it - Add TestDetection_UnseededDestinationDoesNotOverload * fix: handler force_move propagation, partial failure, deterministic dedupe - Propagate ForceMove from outer BalanceTaskParams to individual move TaskParams so batch moves respect the force_move flag - Fix partial failure: mark job successful if at least one move succeeded (succeeded > 0 \|\| failed == 0) to avoid re-running already-completed moves on retry - Use SHA-256 hash for deterministic dedupe key fallback instead of time.Now().UnixNano() which is non-deterministic - Remove unused successDetails variable - Extract maxProposalStringLength constant to replace magic number 200 * admin UI: use template literals in balance execution plan rendering * fix: integration test handles batch proposals from batched detection With batch_size=20, all moves are grouped into a single proposal containing BalanceParams.Moves instead of top-level Sources/Targets. Update assertions to handle both batch and single-move proposal formats. * fix: verify volume size on target before deleting source during balance Add a pre-delete safety check that reads the volume file status on both source and target, then compares .dat file size and file count. If they don't match, the move is aborted — leaving the source intact rather than risking irreversible data loss. Also removes the redundant mountVolume call since VolumeCopy already mounts the volume on the target server. * fix: clamp maxConcurrent, serialize progress sends, validate config as int64 - Clamp maxConcurrentMoves to defaultMaxConcurrentMoves before creating the semaphore so a stale or malicious job cannot request unbounded concurrent volume moves - Extend progressMu to cover sender.SendProgress calls since the underlying gRPC stream is not safe for concurrent writes - Perform bounds checks on max_concurrent_moves and batch_size in int64 space before casting to int, avoiding potential overflow on 32-bit * fix: check disk capacity in resolveBalanceDestination Skip disks where VolumeCount >= MaxVolumeCount so the detection loop does not propose moves to a full disk that would fail at execution time. * test: rename unseeded destination test to match actual behavior The test exercises a server with 0 volumes that IS seeded from topology (matching disk type), not an unseeded destination. Rename to TestDetection_ZeroVolumeServerIncludedInBalance and fix comments. * test: tighten integration test to assert exactly one batch proposal With default batch_size=20, all moves should be grouped into a single batch proposal. Assert len(proposals)==1 and require BalanceParams with Moves, removing the legacy single-move else branch. * fix: propagate ctx to RPCs and restore source writability on abort - All helper methods (markVolumeReadonly, copyVolume, tailVolume, readVolumeFileStatus, deleteVolume) now accept a context parameter instead of using context.Background(), so Execute's ctx propagates cancellation and timeouts into every volume server RPC - Add deferred cleanup that restores the source volume to writable if any step after markVolumeReadonly fails, preventing the source from being left permanently readonly on abort - Add markVolumeWritable helper using VolumeMarkWritableRequest * fix: deep-copy protobuf messages in test recording sender Use proto.Clone in recordingExecutionSender to store immutable snapshots of JobProgressUpdate and JobCompleted, preventing assertions from observing mutations if the handler reuses message pointers. * fix: add VolumeMarkWritable and ReadVolumeFileStatus to fake volume server The balance task now calls ReadVolumeFileStatus for pre-delete verification and VolumeMarkWritable to restore writability on abort. Add both RPCs to the test fake, and drop the mountCalls assertion since BalanceTask no longer calls VolumeMount directly (VolumeCopy handles it). * fix: use maxConcurrentMovesLimit (50) for clamp, not defaultMaxConcurrentMoves defaultMaxConcurrentMoves (5) is the fallback when the field is unset, not an upper bound. Clamping to it silently overrides valid config values like 10/20/50. Introduce maxConcurrentMovesLimit (50) matching the descriptor's MaxValue and clamp to that instead. * fix: cancel batch moves on progress stream failure Derive a cancellable batchCtx from the caller's ctx. If sender.SendProgress returns an error (client disconnect, context cancelled), capture it, skip further sends, and cancel batchCtx so in-flight moves abort via their propagated context rather than running blind to completion. * fix: bound cleanup timeout and validate batch move fields - Use a 30-second timeout for the deferred markVolumeWritable cleanup instead of context.Background() which can block indefinitely if the volume server is unreachable - Validate required fields (VolumeID, SourceNode, TargetNode) before appending moves to a batch proposal, skipping invalid entries - Fall back to a single-move proposal when filtering leaves only one valid move in a batch * fix: cancel task execution on SendProgress stream failure All handler progress callbacks previously ignored SendProgress errors, allowing tasks to continue executing after the client disconnected. Now each handler creates a derived cancellable context and cancels it on the first SendProgress error, stopping the in-flight task promptly. Handlers fixed: erasure_coding, vacuum, volume_balance (single-move), and admin_script (breaks command loop on send failure). * fix: validate batch moves before scheduling in executeBatchMoves Reject empty batches, enforce a hard upper bound (100 moves), and filter out nil or incomplete move specs (missing source/target/volume) before allocating progress tracking and launching goroutines. * test: add batch balance execution integration test Tests the batch move path with 3 volumes, max concurrency 2, using fake volume servers. Verifies all moves complete with correct readonly, copy, tail, and delete RPC counts. * test: add MarkWritableCount and ReadFileStatusCount accessors Expose the markWritableCalls and readFileStatusCalls counters on the fake volume server, following the existing MarkReadonlyCount pattern. * fix: oscillation guard uses global effective counts for heterogeneous capacity The oscillation guard (max-min <= 1) previously used maxServer/minServer which are determined by utilization ratio. With heterogeneous capacity, maxServer by utilization can have fewer raw volumes than minServer, producing a negative diff and incorrectly triggering the guard. Now scans all servers' effective counts to find the true global max/min volume counts, so the guard works correctly regardless of whether utilization-based or raw-count balancing is used. * fix: admin script handler breaks outer loop on SendProgress failure The break on SendProgress error inside the shell.Commands scan only exited the inner loop, letting the outer command loop continue executing commands on a broken stream. Use a sendBroken flag to propagate the break to the outer execCommands loop.	7 days ago
Chris Lu	b991acf634	fix: paginate bucket listing in Admin UI to show all buckets (#8585 ) * fix: paginate bucket listing in Admin UI to show all buckets The Admin UI's GetS3Buckets() had a hardcoded Limit of 1000 in the ListEntries request, causing the Total Buckets count to cap at 1000 even when more buckets exist. This adds pagination to iterate through all buckets by continuing from the last entry name when a full page is returned. Fixes seaweedfs/seaweedfs#8564 * feat: add server-side pagination and sorting to S3 buckets page Add pagination controls, page size selector, and sortable column headers to the Admin UI's Object Store buckets page, following the same pattern used by the Cluster Volumes page. This ensures the UI remains responsive with thousands of buckets. - Add CurrentPage, TotalPages, PageSize, SortBy, SortOrder to S3BucketsData - Accept page/pageSize/sortBy/sortOrder query params in ShowS3Buckets handler - Sort buckets by name, owner, created, objects, logical/physical size - Paginate results server-side (default 100 per page) - Add pagination nav, page size dropdown, and sort indicators to template * Update s3_buckets_templ.go * Update object_store_users_templ.go * fix: use errors.Is(err, io.EOF) instead of string comparison Replace brittle err.Error() == "EOF" string comparison with idiomatic errors.Is(err, io.EOF) for checking stream end in bucket listing. * fix: address PR review findings for bucket pagination - Clamp page to totalPages when page exceeds total, preventing empty results with misleading pagination state - Fix sort comparator to use explicit ascending/descending comparisons with a name tie-breaker, satisfying strict weak ordering for sort.Slice - Capture SnapshotTsNs from first ListEntries response and pass it to subsequent requests for consistent pagination across pages - Replace non-focusable <th onclick> sort headers with <a> tags and reuse getSortIcon, matching the cluster_volumes accessibility pattern - Change exportBucketList() to fetch all buckets from /api/s3/buckets instead of scraping DOM rows (which now only contain the current page)	7 days ago
Chris Lu	02d3e3195c	Update object_store_users_templ.go	7 days ago
Chris Lu	470075dd90	admin/balance: fix Max Volumes display and balancer source selection (#8583 ) * admin: fix Max Volumes column always showing 0 GetClusterVolumeServers() computed DiskCapacity from diskInfo.MaxVolumeCount but never populated the MaxVolumes field on the VolumeServer struct, causing the column to always display 0. * balance: use utilization ratio for source server selection The balancer selected the source server (to move volumes FROM) by raw volume count. In clusters with heterogeneous MaxVolumeCount settings, the server with the highest capacity naturally holds the most volumes and was always picked as the source, even when it had the lowest utilization ratio. Change source selection and imbalance calculation to use utilization ratio (effectiveCount / maxVolumeCount) so servers are compared by how full they are relative to their capacity, not by absolute volume count. This matches how destination scoring already works via calculateBalanceScore().	7 days ago
Lars Lehtonen	f8b7357350	weed/server: fix dropped error (#8584 ) * weed/server: fix dropped error * Removed the redundant check. --------- Co-authored-by: Chris Lu <chrislusf@users.noreply.github.com> Co-authored-by: Chris Lu <chris.lu@gmail.com>	7 days ago
Chris Lu	6dab90472b	admin: fix access key creation UX (#8579 ) * admin: remove misleading "secret key only shown once" warning The access key details modal already allows viewing both the access key and secret key at any time, so the warning about the secret key only being displayed once is incorrect and misleading. * admin: allow specifying custom access key and secret key Add optional access_key and secret_key fields to the create access key API. When provided, the specified keys are used instead of generating random ones. The UI now shows a form with optional fields when creating a new key, with a note that leaving them blank auto-generates keys. * admin: check access key uniqueness before creating Access keys must be globally unique across all users since S3 auth looks them up in a single global map. Add an explicit check using GetUserByAccessKey before creating, so the user gets a clear error ("access key is already in use") rather than a generic store error. * Update object_store_users_templ.go * admin: address review feedback for access key creation Handler: - Use decodeJSONBody/newJSONMaxReader instead of raw json.Decode to enforce request size limits and handle malformed JSON properly - Return 409 Conflict for duplicate access keys, 400 Bad Request for validation errors, instead of generic 500 Backend: - Validate access key length (4-128 chars) and secret key length (8-128 chars) when user-provided Frontend: - Extract resetCreateKeyForm() helper to avoid duplicated cleanup logic - Wire resetCreateKeyForm to accessKeysModal hidden.bs.modal event so form state is always cleared when modal is dismissed - Change secret key input to type="password" with a visibility toggle * admin: guard against nil request and handle GetUserByAccessKey errors - Add nil check for the CreateAccessKeyRequest pointer before dereferencing, defaulting to an empty request (auto-generate both keys). - Handle non-"not found" errors from GetUserByAccessKey explicitly instead of silently proceeding, so store errors (e.g. db connection failures) surface rather than being swallowed. * Update object_store_users_templ.go * admin: fix access key uniqueness check with gRPC store GetUserByAccessKey returns a gRPC NotFound status error (not the sentinel credential.ErrAccessKeyNotFound) when using the gRPC store, causing the uniqueness check to fail with a spurious error. Treat the lookup as best-effort: only reject when a user is found (err == nil). Any error (not-found via any store, connectivity issues) falls through to the store's own CreateAccessKey which enforces uniqueness definitively. * admin: fix error handling and input validation for access key creation Backend: - Remove access key value from the duplicate-key error message to avoid logging the caller-supplied identifier. Handler: - Handle empty POST body (io.EOF) as a valid request that auto-generates both keys, instead of rejecting it as malformed JSON. - Return 404 for "not found" errors (e.g. non-existent user) instead of collapsing them into a 500. Frontend: - Add minlength/maxlength attributes matching backend constraints (access key 4-128, secret key 8-128). - Call reportValidity() before submitting so invalid lengths are caught client-side without a round trip. * admin: use sentinel errors and fix GetUserByAccessKey error handling Backend (user_management.go): - Define sentinel errors (ErrAccessKeyInUse, ErrUserNotFound, ErrInvalidInput) and wrap them in returned errors so callers can use errors.Is. - Handle GetUserByAccessKey errors properly: check the sentinel credential.ErrAccessKeyNotFound first, then fall back to string matching for stores (gRPC) that return non-sentinel not-found errors. Surface unexpected errors instead of silently proceeding. Handler (user_handlers.go): - Replace fragile strings.Contains error matching with errors.Is against the new dash sentinels. Frontend (object_store_users.templ): - Add double-submit guard (isCreatingKey flag + button disabling) to prevent duplicate access key creation requests.	7 days ago
Chris Lu	f8d783f80e	fix: ListObjectVersions interleave Version and DeleteMarker in sort order (#8567 ) * fix: ListObjectVersions interleave Version and DeleteMarker in sort order Go's default xml.Marshal serializes struct fields in definition order, causing all <Version> elements to appear before all <DeleteMarker> elements. The S3 API contract requires these elements to be interleaved in the correct global sort order (by key ascending, then newest version first within each key). This broke clients that validate version list ordering within a single key — an older Version would appear before a newer DeleteMarker for the same object. Fix: Replace the separate Versions/DeleteMarkers/CommonPrefixes arrays with a single Entries []VersionListEntry slice. Each VersionListEntry uses a per-element MarshalXML that outputs the correct XML tag name (<Version>, <DeleteMarker>, or <CommonPrefixes>) based on which field is populated. Since the entries are already in their correct sorted order from buildSortedCombinedList, the XML output is automatically interleaved correctly. Also removes the unused ListObjectVersionsResult struct. Note: The reporter also mentioned a cross-key timestamp ordering issue when paginating with max-keys=1, but that is correct S3 behavior — ListObjectVersions sorts by key name (ascending), not by timestamp. Different keys having non-monotonic timestamps is expected. * test: add CommonPrefixes XML marshaling coverage for ListObjectVersions * fix: validate VersionListEntry has exactly one field set in MarshalXML Return an error instead of silently emitting an empty <Version> element when no field (or multiple fields) are populated. Also clean up the misleading xml:"Version" struct tag on the Entries field.	7 days ago
Chris Lu	55bce53953	reduce logs	7 days ago
Chris Lu	992db11d2b	iam: add IAM group management (#8560 ) * iam: add Group message to protobuf schema Add Group message (name, members, policy_names, disabled) and add groups field to S3ApiConfiguration for IAM group management support (issue #7742). * iam: add group CRUD to CredentialStore interface and all backends Add group management methods (CreateGroup, GetGroup, DeleteGroup, ListGroups, UpdateGroup) to the CredentialStore interface with implementations for memory, filer_etc, postgres, and grpc stores. Wire group loading/saving into filer_etc LoadConfiguration and SaveConfiguration. * iam: add group IAM response types Add XML response types for group management IAM actions: CreateGroup, DeleteGroup, GetGroup, ListGroups, AddUserToGroup, RemoveUserFromGroup, AttachGroupPolicy, DetachGroupPolicy, ListAttachedGroupPolicies, ListGroupsForUser. * iam: add group management handlers to embedded IAM API Add CreateGroup, DeleteGroup, GetGroup, ListGroups, AddUserToGroup, RemoveUserFromGroup, AttachGroupPolicy, DetachGroupPolicy, ListAttachedGroupPolicies, and ListGroupsForUser handlers with dispatch in ExecuteAction. * iam: add group management handlers to standalone IAM API Add group handlers (CreateGroup, DeleteGroup, GetGroup, ListGroups, AddUserToGroup, RemoveUserFromGroup, AttachGroupPolicy, DetachGroupPolicy, ListAttachedGroupPolicies, ListGroupsForUser) and wire into DoActions dispatch. Also add helper functions for user/policy side effects. * iam: integrate group policies into authorization Add groups and userGroups reverse index to IdentityAccessManagement. Populate both maps during ReplaceS3ApiConfiguration and MergeS3ApiConfiguration. Modify evaluateIAMPolicies to evaluate policies from user's enabled groups in addition to user policies. Update VerifyActionPermission to consider group policies when checking hasAttachedPolicies. * iam: add group side effects on user deletion and rename When a user is deleted, remove them from all groups they belong to. When a user is renamed, update group membership references. Applied to both embedded and standalone IAM handlers. * iam: watch /etc/iam/groups directory for config changes Add groups directory to the filer subscription watcher so group file changes trigger IAM configuration reloads. * admin: add group management page to admin UI Add groups page with CRUD operations, member management, policy attachment, and enable/disable toggle. Register routes in admin handlers and add Groups entry to sidebar navigation. * test: add IAM group management integration tests Add comprehensive integration tests for group CRUD, membership, policy attachment, policy enforcement, disabled group behavior, user deletion side effects, and multi-group membership. Add "group" test type to CI matrix in s3-iam-tests workflow. * iam: address PR review comments for group management - Fix XSS vulnerability in groups.templ: replace innerHTML string concatenation with DOM APIs (createElement/textContent) for rendering member and policy lists - Use userGroups reverse index in embedded IAM ListGroupsForUser for O(1) lookup instead of iterating all groups - Add buildUserGroupsIndex helper in standalone IAM handlers; use it in ListGroupsForUser and removeUserFromAllGroups for efficient lookup - Add note about gRPC store load-modify-save race condition limitation * iam: add defensive copies, validation, and XSS fixes for group management - Memory store: clone groups on store/retrieve to prevent mutation - Admin dash: deep copy groups before mutation, validate user/policy exists - HTTP handlers: translate credential errors to proper HTTP status codes, use bool for Enabled field to distinguish missing vs false - Groups templ: use data attributes + event delegation instead of inline onclick for XSS safety, prevent stale async responses iam: add explicit group methods to PropagatingCredentialStore Add CreateGroup, GetGroup, DeleteGroup, ListGroups, and UpdateGroup methods instead of relying on embedded interface fallthrough. Group changes propagate via filer subscription so no RPC propagation needed. * iam: detect postgres unique constraint violation and add groups index Return ErrGroupAlreadyExists when INSERT hits SQLState 23505 instead of a generic error. Add index on groups(disabled) for filtered queries. * iam: add Marker field to group list response types Add Marker string field to GetGroupResult, ListGroupsResult, ListAttachedGroupPoliciesResult, and ListGroupsForUserResult to match AWS IAM pagination response format. * iam: check group attachment before policy deletion Reject DeletePolicy if the policy is attached to any group, matching AWS IAM behavior. Add PolicyArn to ListAttachedGroupPolicies response. * iam: include group policies in IAM authorization Merge policy names from user's enabled groups into the IAMIdentity used for authorization, so group-attached policies are evaluated alongside user-attached policies. * iam: check for name collision before renaming user in UpdateUser Scan identities and inline policies for newUserName before mutating, returning EntityAlreadyExists if a collision is found. Reuse the already-loaded policies instead of loading them again inside the loop. * test: use t.Cleanup for bucket cleanup in group policy test * iam: wrap ErrUserNotInGroup sentinel in RemoveGroupMember error Wrap credential.ErrUserNotInGroup so errors.Is works in groupErrorToHTTPStatus, returning proper 400 instead of 500. * admin: regenerate groups_templ.go with XSS-safe data attributes Regenerated from groups.templ which uses data-group-name attributes instead of inline onclick with string interpolation. * iam: add input validation and persist groups during migration - Validate nil/empty group name in CreateGroup and UpdateGroup - Save groups in migrateToMultiFile so they survive legacy migration * admin: use groupErrorToHTTPStatus in GetGroupMembers and GetGroupPolicies * iam: short-circuit UpdateUser when newUserName equals current name * iam: require empty PolicyNames before group deletion Reject DeleteGroup when group has attached policies, matching the existing members check. Also fix GetGroup error handling in DeletePolicy to only skip ErrGroupNotFound, not all errors. * ci: add weed/pb/** to S3 IAM test trigger paths * test: replace time.Sleep with require.Eventually for propagation waits Use polling with timeout instead of fixed sleeps to reduce flakiness in integration tests waiting for IAM policy propagation. * fix: use credentialManager.GetPolicy for AttachGroupPolicy validation Policies created via CreatePolicy through credentialManager are stored in the credential store, not in s3cfg.Policies (which only has static config policies). Change AttachGroupPolicy to use credentialManager.GetPolicy() for policy existence validation. * feat: add UpdateGroup handler to embedded IAM API Add UpdateGroup action to enable/disable groups and rename groups via the IAM API. This is a SeaweedFS extension (not in AWS SDK) used by tests to toggle group disabled status. * fix: authenticate raw IAM API calls in group tests The embedded IAM endpoint rejects anonymous requests. Replace callIAMAPI with callIAMAPIAuthenticated that uses JWT bearer token authentication via the test framework. * feat: add UpdateGroup handler to standalone IAM API Mirror the embedded IAM UpdateGroup handler in the standalone IAM API for parity. * fix: add omitempty to Marker XML tags in group responses Non-truncated responses should not emit an empty <Marker/> element. * fix: distinguish backend errors from missing policies in AttachGroupPolicy Return ServiceFailure for credential manager errors instead of masking them as NoSuchEntity. Also switch ListGroupsForUser to use s3cfg.Groups instead of in-memory reverse index to avoid stale data. Add duplicate name check to UpdateGroup rename. * fix: standalone IAM AttachGroupPolicy uses persisted policy store Check managed policies from GetPolicies() instead of s3cfg.Policies so dynamically created policies are found. Also add duplicate name check to UpdateGroup rename. * fix: rollback inline policies on UpdateUser PutPolicies failure If PutPolicies fails after moving inline policies to the new username, restore both the identity name and the inline policies map to their original state to avoid a partial-write window. * fix: correct test cleanup ordering for group tests Replace scattered defers with single ordered t.Cleanup in each test to ensure resources are torn down in reverse-creation order: remove membership, detach policies, delete access keys, delete users, delete groups, delete policies. Move bucket cleanup to parent test scope and delete objects before bucket. * fix: move identity nil check before map lookup and refine hasAttachedPolicies Move the nil check on identity before accessing identity.Name to prevent panic. Also refine hasAttachedPolicies to only consider groups that are enabled and have actual policies attached, so membership in a no-policy group doesn't incorrectly trigger IAM authorization. * fix: fail group reload on unreadable or corrupt group files Return errors instead of logging and continuing when group files cannot be read or unmarshaled. This prevents silently applying a partial IAM config with missing group memberships or policies. * fix: use errors.Is for sql.ErrNoRows comparison in postgres group store * docs: explain why group methods skip propagateChange Group changes propagate to S3 servers via filer subscription (watching /etc/iam/groups/) rather than gRPC RPCs, since there are no group-specific RPCs in the S3 cache protocol. * fix: remove unused policyNameFromArn and strings import * fix: update service account ParentUser on user rename When renaming a user via UpdateUser, also update ParentUser references in service accounts to prevent them from becoming orphaned after the next configuration reload. * fix: wrap DetachGroupPolicy error with ErrPolicyNotAttached sentinel Use credential.ErrPolicyNotAttached so groupErrorToHTTPStatus maps it to 400 instead of falling back to 500. * fix: use admin S3 client for bucket cleanup in enforcement test The user S3 client may lack permissions by cleanup time since the user is removed from the group in an earlier subtest. Use the admin S3 client to ensure bucket and object cleanup always succeeds. * fix: add nil guard for group param in propagating store log calls Prevent potential nil dereference when logging group.Name in CreateGroup and UpdateGroup of PropagatingCredentialStore. * fix: validate Disabled field in UpdateGroup handlers Reject values other than "true" or "false" with InvalidInputException instead of silently treating them as false. * fix: seed mergedGroups from existing groups in MergeS3ApiConfiguration Previously the merge started with empty group maps, dropping any static-file groups. Now seeds from existing iam.groups before overlaying dynamic config, and builds the reverse index after merging to avoid stale entries from overridden groups. * fix: use errors.Is for filer_pb.ErrNotFound comparison in group loading Replace direct equality (==) with errors.Is() to correctly match wrapped errors, consistent with the rest of the codebase. * fix: add ErrUserNotFound and ErrPolicyNotFound to groupErrorToHTTPStatus Map these sentinel errors to 404 so AddGroupMember and AttachGroupPolicy return proper HTTP status codes. * fix: log cleanup errors in group integration tests Replace fire-and-forget cleanup calls with error-checked versions that log failures via t.Logf for debugging visibility. * fix: prevent duplicate group test runs in CI matrix The basic lane's -run "TestIAM" regex also matched TestIAMGroup* tests, causing them to run in both the basic and group lanes. Replace with explicit test function names. * fix: add GIN index on groups.members JSONB for membership lookups Without this index, ListGroupsForUser and membership queries require full table scans on the groups table. * fix: handle cross-directory moves in IAM config subscription When a file is moved out of an IAM directory (e.g., /etc/iam/groups), the dir variable was overwritten with NewParentPath, causing the source directory change to be missed. Now also notifies handlers about the source directory for cross-directory moves. * fix: validate members/policies before deleting group in admin handler AdminServer.DeleteGroup now checks for attached members and policies before delegating to credentialManager, matching the IAM handler guards. * fix: merge groups by name instead of blind append during filer load Match the identity loader's merge behavior: find existing group by name and replace, only append when no match exists. Prevents duplicates when legacy and multi-file configs overlap. * fix: check DeleteEntry response error when cleaning obsolete group files Capture and log resp.Error from filer DeleteEntry calls during group file cleanup, matching the pattern used in deleteGroupFile. * fix: verify source user exists before no-op check in UpdateUser Reorder UpdateUser to find the source identity first and return NoSuchEntityException if not found, before checking if the rename is a no-op. Previously a non-existent user renamed to itself would incorrectly return success. * fix: update service account parent refs on user rename in embedded IAM The embedded IAM UpdateUser handler updated group membership but not service account ParentUser fields, unlike the standalone handler. * fix: replay source-side events for all handlers on cross-dir moves Pass nil newEntry to bucket, IAM, and circuit-breaker handlers for the source directory during cross-directory moves, so all watchers can clear caches for the moved-away resource. * fix: don't seed mergedGroups from existing iam.groups in merge Groups are always dynamic (from filer), never static (from s3.config). Seeding from iam.groups caused stale deleted groups to persist. Now only uses config.Groups from the dynamic filer config. * fix: add deferred user cleanup in TestIAMGroupUserDeletionSideEffect Register t.Cleanup for the created user so it gets cleaned up even if the test fails before the inline DeleteUser call. * fix: assert UpdateGroup HTTP status in disabled group tests Add require.Equal checks for 200 status after UpdateGroup calls so the test fails immediately on API errors rather than relying on the subsequent Eventually timeout. * fix: trim whitespace from group name in filer store operations Trim leading/trailing whitespace from group.Name before validation in CreateGroup and UpdateGroup to prevent whitespace-only filenames. Also merge groups by name during multi-file load to prevent duplicates. * fix: add nil/empty group validation in gRPC store Guard CreateGroup and UpdateGroup against nil group or empty name to prevent panics and invalid persistence. * fix: add nil/empty group validation in postgres store Guard CreateGroup and UpdateGroup against nil group or empty name to prevent panics from nil member access and empty-name row inserts. * fix: add name collision check in embedded IAM UpdateUser The embedded IAM handler renamed users without checking if the target name already existed, unlike the standalone handler. * fix: add ErrGroupNotEmpty sentinel and map to HTTP 409 AdminServer.DeleteGroup now wraps conflict errors with ErrGroupNotEmpty, and groupErrorToHTTPStatus maps it to 409 Conflict instead of 500. * fix: use appropriate error message in GetGroupDetails based on status Return "Group not found" only for 404, use "Failed to retrieve group" for other error statuses instead of always saying "Group not found". * fix: use backend-normalized group.Name in CreateGroup response After credentialManager.CreateGroup may normalize the name (e.g., trim whitespace), use group.Name instead of the raw input for the returned GroupData to ensure consistency. * fix: add nil/empty group validation in memory store Guard CreateGroup and UpdateGroup against nil group or empty name to prevent panics from nil pointer dereference on map access. * fix: reorder embedded IAM UpdateUser to verify source first Find the source identity before checking for collisions, matching the standalone handler's logic. Previously a non-existent user renamed to an existing name would get EntityAlreadyExists instead of NoSuchEntity. * fix: handle same-directory renames in metadata subscription Replay a delete event for the old entry name during same-directory renames so handlers like onBucketMetadataChange can clean up stale state for the old name. * fix: abort GetGroups on non-ErrGroupNotFound errors Only skip groups that return ErrGroupNotFound. Other errors (e.g., transient backend failures) now abort the handler and return the error to the caller instead of silently producing partial results. * fix: add aria-label and title to icon-only group action buttons Add accessible labels to View and Delete buttons so screen readers and tooltips provide meaningful context. * fix: validate group name in saveGroup to prevent invalid filenames Trim whitespace and reject empty names before writing group JSON files, preventing creation of files like ".json". * fix: add /etc/iam/groups to filer subscription watched directories The groups directory was missing from the watched directories list, so S3 servers in a cluster would not detect group changes made by other servers via filer. The onIamConfigChange handler already had code to handle group directory changes but it was never triggered. * add direct gRPC propagation for group changes to S3 servers Groups now have the same dual propagation as identities and policies: direct gRPC push via propagateChange + async filer subscription. - Add PutGroup/RemoveGroup proto messages and RPCs - Add PutGroup/RemoveGroup in-memory cache methods on IAM - Add PutGroup/RemoveGroup gRPC server handlers - Update PropagatingCredentialStore to call propagateChange on group mutations * reduce log verbosity for config load summary Change ReplaceS3ApiConfiguration log from Infof to V(1).Infof to avoid noisy output on every config reload. * admin: show user groups in view and edit user modals - Add Groups field to UserDetails and populate from credential manager - Show groups as badges in user details view modal - Add group management to edit user modal: display current groups, add to group via dropdown, remove from group via badge x button * fix: remove duplicate showAlert that broke modal-alerts.js admin.js defined showAlert(type, message) which overwrote the modal-alerts.js version showAlert(message, type), causing broken unstyled alert boxes. Remove the duplicate and swap all callers in admin.js to use the correct (message, type) argument order. * fix: unwrap groups API response in edit user modal The /api/groups endpoint returns {"groups": [...]}, not a bare array. * Update object_store_users_templ.go * test: assert AccessDenied error code in group denial tests Replace plain assert.Error checks with awserr.Error type assertion and AccessDenied code verification, matching the pattern used in other IAM integration tests. * fix: propagate GetGroups errors in ShowGroups handler getGroupsPageData was swallowing errors and returning an empty page with 200 status. Now returns the error so ShowGroups can respond with a proper error status. * fix: reject AttachGroupPolicy when credential manager is nil Previously skipped policy existence validation when credentialManager was nil, allowing attachment of nonexistent policies. Now returns a ServiceFailureException error. * fix: preserve groups during partial MergeS3ApiConfiguration updates UpsertIdentity calls MergeS3ApiConfiguration with a partial config containing only the updated identity (nil Groups). This was wiping all in-memory group state. Now only replaces groups when config.Groups is non-nil (full config reload). * fix: propagate errors from group lookup in GetObjectStoreUserDetails ListGroups and GetGroup errors were silently ignored, potentially showing incomplete group data in the UI. * fix: use DOM APIs for group badge remove button to prevent XSS Replace innerHTML with onclick string interpolation with DOM createElement + addEventListener pattern. Also add aria-label and title to the add-to-group button. * fix: snapshot group policies under RLock to prevent concurrent map access evaluateIAMPolicies was copying the map reference via groupMap := iam.groups under RLock then iterating after RUnlock, while PutGroup mutates the map in-place. Now copies the needed policy names into a slice while holding the lock. * fix: add nil IAM check to PutGroup and RemoveGroup gRPC handlers Match the nil guard pattern used by PutPolicy/DeletePolicy to prevent nil pointer dereference when IAM is not initialized.	7 days ago
Chris Lu	78a3441b30	fix: volume balance detection returns multiple tasks per run (#8559 ) * fix: volume balance detection now returns multiple tasks per run (#8551) Previously, detectForDiskType() returned at most 1 balance task per disk type, making the MaxJobsPerDetection setting ineffective. The detection loop now iterates within each disk type, planning multiple moves until the imbalance drops below threshold or maxResults is reached. Effective volume counts are adjusted after each planned move so the algorithm correctly re-evaluates which server is overloaded. * fix: factor pending tasks into destination scoring and use UnixNano for task IDs - Use UnixNano instead of Unix for task IDs to avoid collisions when multiple tasks are created within the same second - Adjust calculateBalanceScore to include LoadCount (pending + assigned tasks) in the utilization estimate, so the destination picker avoids stacking multiple planned moves onto the same target disk * test: add comprehensive balance detection tests for complex scenarios Cover multi-server convergence, max-server shifting, destination spreading, pre-existing pending task skipping, no-duplicate-volume invariant, and parameterized convergence verification across different cluster shapes and thresholds. * fix: address PR review findings in balance detection - hasMore flag: compute from len(results) >= maxResults so the scheduler knows more pages may exist, matching vacuum/EC handler pattern - Exhausted server fallthrough: when no eligible volumes remain on the current maxServer (all have pending tasks) or destination planning fails, mark the server as exhausted and continue to the next overloaded server instead of stopping the entire detection loop - Return canonical destination server ID directly from createBalanceTask instead of resolving via findServerIDByAddress, eliminating the fragile address→ID lookup for adjustment tracking - Fix bestScore sentinel: use math.Inf(-1) instead of -1.0 so disks with negative scores (high pending load, same rack/DC) are still selected as the best available destination - Add TestDetection_ExhaustedServerFallsThrough covering the scenario where the top server's volumes are all blocked by pre-existing tasks * test: fix computeEffectiveCounts and add len guard in no-duplicate test - computeEffectiveCounts now takes a servers slice to seed counts for all known servers (including empty ones) and uses an address→ID map from the topology spec instead of scanning metrics, so destination servers with zero initial volumes are tracked correctly - TestDetection_NoDuplicateVolumesAcrossIterations now asserts len > 1 before checking duplicates, so the test actually fails if Detection regresses to returning a single task * fix: remove redundant HasAnyTask check in createBalanceTask The HasAnyTask check in createBalanceTask duplicated the same check already performed in detectForDiskType's volume selection loop. Since detection runs single-threaded (MaxDetectionConcurrency: 1), no race can occur between the two points. * fix: consistent hasMore pattern and remove double-counted LoadCount in scoring - Adopt vacuum_handler's hasMore pattern: over-fetch by 1, check len > maxResults, and truncate — consistent truncation semantics - Remove direct LoadCount penalty in calculateBalanceScore since LoadCount is already factored into effectiveVolumeCount for utilization scoring; bump utilization weight from 40 to 50 to compensate for the removed 10-point load penalty * fix: handle zero maxResults as no-cap, emit trace after trim, seed empty servers - When MaxResults is 0 (omitted), treat as no explicit cap instead of defaulting to 1; only apply the +1 over-fetch probe when caller supplies a positive limit - Move decision trace emission after hasMore/trim so the trace accurately reflects the returned proposals - Seed serverVolumeCounts from ActiveTopology so servers that have a matching disk type but zero volumes are included in the imbalance calculation and MinServerCount check * fix: nil-guard clusterInfo, uncap legacy DetectionFunc, deterministic disk type order - Add early nil guard for clusterInfo in Detection to prevent panics in downstream helpers (detectForDiskType, createBalanceTask) - Change register.go DetectionFunc wrapper from maxResults=1 to 0 (no cap) so the legacy code path returns all detected tasks - Sort disk type keys before iteration so results are deterministic when maxResults spans multiple disk types (HDD/SSD) * fix: don't over-fetch in stateful detection to avoid orphaned pending tasks Detection registers planned moves in ActiveTopology via AddPendingTask, so requesting maxResults+1 would create an extra pending task that gets discarded during trim. Use len(results) >= maxResults as the hasMore signal instead, which is correct since Detection already caps internally. * fix: return explicit truncated flag from Detection instead of approximating Detection now returns (results, truncated, error) where truncated is true only when the loop stopped because it hit maxResults, not when it ran out of work naturally. This eliminates false hasMore signals when detection happens to produce exactly maxResults results by resolving the imbalance. * cleanup: simplify detection logic and remove redundancies - Remove redundant clusterInfo nil check in detectForDiskType since Detection already guards against nil clusterInfo - Remove adjustments loop for destination servers not in serverVolumeCounts — topology seeding ensures all servers with matching disk type are already present - Merge two-loop min/max calculation into a single loop: min across all servers, max only among non-exhausted servers - Replace magic number 100 with len(metrics) for minC initialization in convergence test * fix: accurate truncation flag, deterministic server order, indexed volume lookup - Track balanced flag to distinguish "hit maxResults cap" from "cluster balanced at exactly maxResults" — truncated is only true when there's genuinely more work to do - Sort servers for deterministic iteration and tie-breaking when multiple servers have equal volume counts - Pre-index volumes by server with per-server cursors to avoid O(maxResults * volumes) rescanning on each iteration - Add truncation flag assertions to RespectsMaxResults test: true when capped, false when detection finishes naturally * fix: seed trace server counts from ActiveTopology to match detection logic The decision trace was building serverVolumeCounts only from metrics, missing zero-volume servers seeded from ActiveTopology by Detection. This could cause the trace to report wrong server counts, incorrect imbalance ratios, or spurious "too few servers" messages. Pass activeTopology into the trace function and seed server counts the same way Detection does. * fix: don't exhaust server on per-volume planning failure, sort volumes by ID - When createBalanceTask returns nil, continue to the next volume on the same server instead of marking the entire server as exhausted. The failure may be volume-specific (not found in topology, pending task registration failed) and other volumes on the server may still be viable candidates. - Sort each server's volume slice by VolumeID after pre-indexing so volume selection is fully deterministic regardless of input order. * fix: use require instead of assert to prevent nil dereference panic in CORS test The test used assert.NoError (non-fatal) for GetBucketCors, then immediately accessed getResp.CORSRules. When the API returns an error, getResp is nil causing a panic. Switch to require.NoError/NotNil/Len so the test stops before dereferencing a nil response. * fix: deterministic disk tie-breaking and stronger pre-existing task test - Sort available disks by NodeID then DiskID before scoring so destination selection is deterministic when two disks score equally - Add task count bounds assertion to SkipsPreExistingPendingTasks test: with 15 of 20 volumes already having pending tasks, at most 5 new tasks should be created and at least 1 (imbalance still exists) * fix: seed adjustments from existing pending/assigned tasks to prevent over-scheduling Detection now calls ActiveTopology.GetTaskServerAdjustments() to initialize the adjustments map with source/destination deltas from existing pending and assigned balance tasks. This ensures effectiveCounts reflects in-flight moves, preventing the algorithm from planning additional moves in the same direction when prior moves already address the imbalance. Added GetTaskServerAdjustments(taskType) to ActiveTopology which iterates pending and assigned tasks, decrementing source servers and incrementing destination servers for the given task type.	1 week ago
Chris Lu	2ec0a67ee3	master: return 503/Unavailable during topology warmup after leader change (#8529 ) * master: return 503/Unavailable during topology warmup after leader change After a master restart or leader change, the topology is empty until volume servers reconnect and send heartbeats. During this warmup window (3 heartbeat intervals = 15 seconds), volume lookups that fail now return 503 Service Unavailable (HTTP) or gRPC Unavailable instead of 404 Not Found, signaling clients to retry with other masters. * master: skip warmup 503 on fresh start and single-master setups - Check MaxVolumeId > 0 to distinguish restart from fresh start (MaxVolumeId is Raft-persisted, so 0 means no prior data) - Check peer count > 1 so single-master deployments aren't affected (no point suggesting "retry with other masters" if there are none) * master: address review feedback and block assigns during warmup - Protect LastLeaderChangeTime with dedicated mutex (fix data race) - Extract warmup multiplier as WarmupPulseMultiplier constant - Derive Retry-After header from pulse config instead of hardcoding - Only trigger warmup 503 for "not found" errors, not parse errors - Return nil response (not partial) on gRPC Unavailable - Add doc comments to IsWarmingUp, getter/setter, WarmupDuration - Block volume assign requests (HTTP and gRPC) during warmup, since the topology is incomplete and assignments would be unreliable - Skip warmup behavior for single-master setups (no peers to retry) * master: apply warmup to all setups, skip only on fresh start Single-master restarts still have an empty topology until heartbeats arrive, so warmup protection should apply there too. The only case to skip is a fresh cluster start (MaxVolumeId == 0), which already has no volumes to look up. - Remove GetMasterCount() > 1 guard from all warmup checks - Remove now-unused GetMasterCount helper - Update error messages to "topology is still loading" (not "retry with other masters" which doesn't apply to single-master) * master: add client-side retry on Unavailable for lookup and assign The server-side 503/Unavailable during warmup needs client cooperation. Previously, LookupVolumeIds and Assign would immediately propagate the error without retry. Now both paths retry with exponential backoff (1s -> 1.5s -> ... up to 6s) when receiving Unavailable, respecting context cancellation. This covers the warmup window where the master's topology is still loading after a restart or leader change. * master: seed warmup timestamp in legacy raft path at setup The legacy raft path only set lastLeaderChangeTime inside the event listener callback, which could fire after IsLeader() was already observed as true in SetRaftServer. Seed the timestamp at setup time (matching the hashicorp path) so IsWarmingUp() is active immediately. * master: fix assign retry loop to cover full warmup window The retry loop used waitTime <= maxWaitTime as a stop condition, causing it to give up after ~13s while warmup lasts 15s. Now cap each individual sleep at maxWaitTime but keep retrying until the context is cancelled. * master: preserve gRPC status in lookup retry and fix retry window Return the raw gRPC error instead of wrapping with fmt.Errorf so status.FromError() can extract the status code. Use proper gRPC status check (codes.Unavailable) instead of string matching. Also cap individual sleep at maxWaitTime while retrying until ctx is done. * master: use gRPC status code instead of string matching in assign retry Use status.FromError/codes.Unavailable instead of brittle strings.Contains for detecting retriable gRPC errors in the assign retry loop. * master: use remaining warmup duration for Retry-After header Set Retry-After to the remaining warmup time instead of the full warmup duration, so clients don't wait longer than necessary. * master: reset ret.Replicas before populating from assign response Clear Replicas slice before appending to prevent duplicate entries when the assign response is retried or when alternative requests are attempted. * master: add unit tests for warmup retry behavior Test that Assign() and LookupVolumeIds() retry on codes.Unavailable and stop promptly when the context is cancelled. * master: record leader change time before initialization work Move SetLastLeaderChangeTime() to fire immediately when the leader change event is received, before DoBarrier(), EnsureTopologyId(), and updatePeers(), so the warmup clock starts at the true moment of leadership transition. * master: use topology warmup duration in volume growth wait loop Replace hardcoded constants.VolumePulsePeriod * 2 with topo.IsWarmingUp() and topo.WarmupDuration() so the growth wait stays in sync with the configured warmup window. Remove unused constants import. * master: resolve master before creating RPC timeout context Move GetMaster() call before context.WithTimeout() so master resolution blocking doesn't consume the gRPC call timeout. * master: use NotFound flag instead of string matching for volume lookup Add a NotFound field to LookupResult and set it in findVolumeLocation when a volume is genuinely missing. Update HTTP and gRPC warmup checks to use this flag instead of strings.Contains on the error message. * master: bound assign retry loop to 30s for deadline-free contexts Without a context deadline, the Unavailable retry loop could spin forever. Add a maxRetryDuration of 30s so the loop gives up even when no context deadline is set. * master: strengthen assign retry cancellation test Verify the retry loop actually retried (callCount > 1) and that the returned error is context.DeadlineExceeded, not just any error. * master: extract shared retry-with-backoff utility Add util.RetryWithBackoff for context-aware, bounded retry with exponential backoff. Refactor both Assign() and LookupVolumeIds() to use it instead of duplicating the retry/sleep/backoff logic. * master: cap waitTime in RetryWithBackoff to prevent unbounded growth Cap the backoff waitTime at maxWaitTime so it doesn't grow indefinitely in long-running retry scenarios. * master: only return Unavailable during warmup when all lookups failed For batched LookupVolume requests, return partial results when some volumes are found. Only return codes.Unavailable when no volumes were successfully resolved, so clients benefit from partial results instead of retrying unnecessarily. * master: set retriable error message in 503 response body When returning 503 during warmup, replace the "not found" error in the JSON body with "service warming up, please retry" so clients don't treat it as a permanent error. * master: guard empty master address in LookupVolumeIds If GetMaster() returns empty (no master found or ctx cancelled), return an appropriate error instead of dialing an empty address. Returns ctx.Err() if context is done, otherwise codes.Unavailable to trigger retry. * master: add comprehensive tests for RetryWithBackoff Test success after retries, non-retryable error handling, context cancellation, and maxDuration cap with context.Background(). * master: enforce hard maxDuration bound in RetryWithBackoff Use a deadline instead of elapsed-time check so the last sleep is capped to remaining time. This prevents the total retry duration from overshooting maxDuration by up to one full backoff interval. * master: respect fresh-start bypass in RemainingWarmupDuration Check IsWarmingUp() first (which returns false when MaxVolumeId==0) so RemainingWarmupDuration returns 0 on fresh clusters. * master: round up Retry-After seconds to avoid underestimating Use math.Ceil so fractional remaining seconds (e.g. 1.9s) round up to the next integer (2) instead of flooring down (1). * master: tighten batch lookup warmup to all-NotFound only Only return codes.Unavailable when every requested volume ID was a transient not-found. Mixed cases with non-NotFound errors now return the response with per-volume error details preserved. * master: reduce retry log noise and fix timer leak Lower per-attempt retry log from V(0) to V(1) to reduce noise during warmup. Replace time.After with time.NewTimer to avoid lingering timers when context is cancelled. * master: add per-attempt timeout for assign RPC Use a 10s per-attempt timeout so a single slow RPC can't consume the entire 30s retry budget when ctx has no deadline. * master: share single 30s retry deadline across assign request entries The Assign() function iterates over primary and fallback requests, previously giving each its own 30s RetryWithBackoff budget. With a primary + fallback, the total could reach 60s. Compute one deadline up front and pass the remaining budget to each RetryWithBackoff call so the entire Assign() call stays within a single 30s cap. * master: strengthen context-cancel test with DeadlineExceeded and retry assertions Assert errors.Is(err, context.DeadlineExceeded) to verify the error is specifically from the context deadline, and check callCount > 1 to prove retries actually occurred before cancellation. Mirrors the pattern used in TestAssignStopsOnContextCancel. * master: bound GetMaster with per-attempt timeout in LookupVolumeIds GetMaster() calls WaitUntilConnected() which can block indefinitely if no master is available. Previously it used the outer ctx, so a slow master resolution could consume the entire RetryWithBackoff budget in a single attempt. Move the per-attempt timeoutCtx creation before the GetMaster call so both master resolution and the gRPC LookupVolume RPC share one grpcTimeout-bounded attempt. * master: use deadline-aware context for assign retry budget The shared 30s deadline only limited RetryWithBackoff's internal wall-clock tracking, but per-attempt contexts were still derived from the original ctx and could run for up to 10s even when the budget was nearly exhausted. Create a deadlineCtx from the computed deadline and derive both RetryWithBackoff and per-attempt timeouts from it so all operations honor the shared 30s cap. * master: skip warmup gate for empty lookup requests When VolumeOrFileIds is empty, notFoundCount == len(req.VolumeOrFileIds) is 0 == 0 which is true, causing empty lookup batches during warmup to return codes.Unavailable and be retried endlessly. Add a len(req.VolumeOrFileIds) > 0 guard so empty requests pass through. * master: validate request fields before warmup gate in Assign Move Replication and Ttl parsing before the IsWarmingUp() check so invalid inputs get a proper validation error instead of being masked by codes.Unavailable during warmup. Pure syntactic validation does not depend on topology state and should run first. * master: check deadline and context before starting retry attempt RetryWithBackoff only checked the deadline and context after an attempt completed or during the sleep select. If the deadline expired or context was canceled during sleep, the next iteration would still call operation() before detecting it. Add pre-operation checks so no new attempt starts after the budget is exhausted. * master: always return ctx.Err() on context cancellation in RetryWithBackoff When ctx.Err() is non-nil, the pre-operation check was returning lastErr instead of ctx.Err(). This broke callers checking errors.Is(err, context.DeadlineExceeded) and contradicted the documented contract. Always return ctx.Err() so the cancellation reason is properly surfaced. * master: handle warmup errors in StreamAssign without killing the stream StreamAssign was returning codes.Unavailable errors from Assign directly, which terminates the gRPC stream and breaks pooled connections. Instead, return transient errors as in-band error responses so the stream survives warmup periods. Also reset assignClient in doAssign on Send/Recv failures so a broken stream doesn't leave the proxy permanently dead. * master: wait for warmup before slot search in findAndGrow findEmptySlotsForOneVolume was called before the warmup wait loop, selecting slots from an incomplete topology. Move the warmup wait before slot search so volume placement uses the fully warmed-up topology with all servers registered. * master: add Retry-After header to /dir/assign warmup response The /dir/lookup handler already sets Retry-After during warmup but /dir/assign did not, leaving HTTP clients without guidance on when to retry. Add the same header using RemainingWarmupDuration(). * master: only seed warmup timestamp on leader at startup SetLastLeaderChangeTime was called unconditionally for both leader and follower nodes. Followers don't need warmup state, and the leader change event listener handles real elections. Move the seed into the IsLeader() block so only the startup leader gets warmup initialized. * master: preserve codes.Unavailable for StreamAssign warmup errors in doAssign StreamAssign returns transient warmup errors as in-band AssignResponse.Error messages. doAssign was converting these to plain fmt.Errorf, losing the codes.Unavailable classification needed for the caller's retry logic. Detect warmup error messages and wrap them as status.Error(codes.Unavailable) so RetryWithBackoff can retry.	1 week ago
Chris Lu	0647f66bb5	filer.sync: add exponential backoff on unexpected EOF during replication (#8557 ) * filer.sync: add exponential backoff on unexpected EOF during replication When the source volume server drops connections under high traffic, filer.sync retries aggressively (every 1-6s), hammering the already overloaded source. This adds a longer exponential backoff (10s to 2min) specifically for "unexpected EOF" errors, reducing pressure on the source while still retrying indefinitely until success. Also adds more logging throughout the replication path: - Log source URL and error at V(0) when ReadPart or io.ReadAll fails - Log content-length and byte counts at V(4) on success - Log backoff duration in retry messages Fixes #8542 * filer.sync: extract backoff helper and fix 2-minute cap - Extract nextEofBackoff() and isEofError() helpers to deduplicate the backoff logic between fetchAndWrite and uploadManifestChunk - Fix the cap: previously 80s would double to 160s and pass the < 2min check uncapped. Now doubles first, then clamps to 2min. * filer.sync: log source URL instead of empty upload URL on read errors UploadUrl is not populated until after the reader is consumed, so the V(0) and V(4) logs were printing an empty string. Add SourceUrl field to UploadOption and populate it from the HTTP response in fetchAndWrite. * filer.sync: guard isEofError against nil error * filer.sync: use errors.Is for EOF detection, fix log wording - Replace broad substring matching ("read input", "unexpected EOF") with errors.Is(err, io.ErrUnexpectedEOF) and errors.Is(err, io.EOF) so only actual EOF errors trigger the longer backoff - Fix awkward log phrasing: "interrupted replicate" → "interrupted while replicating" * filer.sync: remove EOF backoff from uploadManifestChunk uploadManifestChunk reads from an in-memory bytes.Reader, so any EOF errors there are from the destination side, not a broken source stream. The long source-oriented backoff is inappropriate; let RetryUntil handle destination retries at its normal cadence. --------- Co-authored-by: Copilot <copilot@github.com>	1 week ago
Chris Lu	ba66411337	Update plugin_templ.go	1 week ago
Chris Lu	7808b301ef	admin: remove Scheduler Settings cards from plugin UI (#8558 ) * admin: remove Scheduler Settings cards, make Next Run full-width Remove the two "Scheduler Settings" placeholder cards from the plugin UI (overview page and scheduler tab). They only contained a text note saying detection intervals are configured per job type, which is self-evident from the per-job-type settings form. Make the "Next Run" card full-width on the overview page since it no longer shares a row with the removed card. * plugin UI: promote Next Run to top summary card row Move "Next Run" from a standalone card into the top row alongside Workers, Active Jobs, and Activities as a compact stat card.	1 week ago
Chris Lu	fa7da0f57e	template	1 week ago
Chris Lu	961c270aba	admin: expose per-job-type detection interval in plugin UI (#8552 ) * admin: expose per-job-type detection interval in plugin UI The detection_interval_seconds field was not editable in the admin UI. collectAdminSettings() silently preserved the existing value, making it impossible for users to change how often a job type checks for new work. Users would change the global "Sleep Between Iterations" setting expecting it to control job scheduling frequency, but that only controls the scheduler loop's idle polling rate. Add a "Detection Interval (s)" input to the per-job-type admin settings form so users can actually configure it. Fixes #8549 * admin: remove global Sleep Between Iterations setting Now that per-job-type detection intervals are exposed in the UI, the global IdleSleepSeconds setting is redundant and confusing. It only controlled the scheduler loop's idle polling rate, which is always overridden by earliestNextDetectionAt() when job types exist. Replace the three usages with simpler alternatives: - Scheduler loop sleep: use defaultSchedulerIdleSleep constant - Initial delay for new job types: use policy.DetectionInterval/2 (more logical since it's already per-job-type) - Status fallback: use the constant The API endpoints are kept for backward compatibility but the UI no longer exposes or calls them. * admin: restore configurable idle sleep in scheduler loop The EC integration test sets idle_sleep_seconds=1 via the scheduler config API so the scheduler wakes quickly after workers connect. The previous commit replaced this with a hardcoded 613s constant, causing the scheduler to sleep through the entire test window. Restore GetSchedulerConfig().IdleSleepDuration() in the scheduler loop and status reporting. The UI removal of the setting is still correct — the API endpoint remains for programmatic use (e.g., tests). * admin: cap first-run initial delay to 5s instead of DetectionInterval/2 The initial delay for first-run job types was set to policy.DetectionInterval/2, which creates unbounded first-run latency (e.g., 1 hour for vacuum with a 2-hour detection interval). A small fixed 5-second delay provides sufficient stagger without penalizing startup time.	1 week ago
Chris Lu	e25558e4d8	admin: fix mobile sidebar menu inaccessible in portrait mode (#8556 ) * admin: fix mobile sidebar menu inaccessible in portrait mode The hamburger button only toggled the user dropdown, leaving the sidebar navigation inaccessible on mobile devices in portrait mode. Add a dedicated sidebar toggle button (visible only on mobile), give the sidebar an id so Bootstrap collapse can target it, add a backdrop overlay for the open state, and auto-close the sidebar when a nav link is clicked. Fixes #8550 * admin: address review feedback on mobile sidebar - Remove redundant JS show/hide.bs.collapse listeners; CSS sibling selector already handles backdrop visibility - Use const instead of var for non-reassigned variables - Move inline style on user icon to CSS class * admin: add aria attributes to user-menu toggler, use CSS variable for navbar height - Add aria-controls, aria-expanded, and aria-label to the user-menu toggle button for assistive technology - Extract hard-coded 56px navbar height into --navbar-height CSS custom property used by sidebar and backdrop positioning * admin: extract hideSidebar helper, use toggler visibility for breakpoint check - Extract duplicated collapse-hide logic into a hideSidebar helper - Replace hardcoded window.innerWidth < 768 with a check on the sidebar toggler's computed display, decoupling JS from CSS breakpoints - Add aria-expanded="false" to sidebar toggle button --------- Co-authored-by: Copilot <copilot@github.com>	1 week ago
Chris Lu	587c24ec89	plugin worker: support job type categories (all, default, heavy) (#8547 ) * plugin worker: add handler registry with job categories Introduce a self-registration pattern for plugin worker job handlers. Each handler can register itself via init() with a HandlerFactory that declares its job type, category (default/heavy), CLI aliases, and a builder function. ResolveHandlerFactories accepts a mix of category names ("all", "default", "heavy") and explicit job type names/aliases, returning the matching factories. This enables workers to be configured by resource profile rather than requiring explicit job type enumeration. * plugin worker: register all handlers via init() Each job handler now self-registers into the global handler registry with its canonical job type, category, CLI aliases, and build function: - vacuum: category=default - volume_balance: category=default - admin_script: category=default - erasure_coding: category=heavy - iceberg_maintenance: category=heavy Adding a new job type now only requires adding the init() call in the handler file itself — no other files need to be touched. * plugin worker: replace hardcoded job type switch with registry Remove buildPluginWorkerHandler, parsePluginWorkerJobTypes, and canonicalPluginWorkerJobType from worker_runtime.go. The simplified buildPluginWorkerHandlers now delegates to pluginworker.ResolveHandlerFactories, which resolves category names ("all", "default", "heavy") and explicit job type names/aliases. The default job type is changed from an explicit list to "all", so new handlers registered via init() are automatically picked up. Update all tests to use the new API. * plugin worker: update CLI help text for job categories Update the -jobType flag description and command examples to document category support (all, default, heavy) alongside explicit job type names. * plugin worker: address review feedback - Add CategoryAll constant; use typed constants in tokenAsCategory - Pre-allocate result slice in ResolveHandlerFactories - Add vacuum aliases (vol.vacuum, volume.vacuum) - List alias examples (ec, balance, iceberg) in -jobType flag help - Create handlers aggregator package for subpackage blank imports so new handler subpackages only need to be added in one place - Make category tests relationship-based (subset/union checks) instead of asserting exact handler counts - Add clarifying comments to worker_test.go and mini_plugin_test.go listing expected handler names next to count assertions --------- Co-authored-by: Copilot <copilot@github.com>	1 week ago
Chris Lu	72c2c7ef8b	Add iceberg_maintenance plugin worker handler (Phase 1) (#8501 ) * Add iceberg_maintenance plugin worker handler (Phase 1) Implement automated Iceberg table maintenance as a new plugin worker job type. The handler scans S3 table buckets for tables needing maintenance and executes operations in the correct Iceberg order: expire snapshots, remove orphan files, and rewrite manifests. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Fix unsafe int64→int narrowing for MaxSnapshotsToKeep Use int64(wouldKeep) instead of int(config.MaxSnapshotsToKeep) to avoid potential truncation on 32-bit platforms (CodeQL high severity). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Fix unsafe int64→int narrowing for MinInputFiles Use int64(len(manifests)) instead of int(config.MinInputFiles) to avoid potential truncation on 32-bit platforms (CodeQL high severity). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Fix unsafe int64→int narrowing for MaxCommitRetries Clamp MaxCommitRetries to [1,20] range and keep as int64 throughout the retry loop to avoid truncation on 32-bit platforms (CodeQL high severity). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Sort snapshots explicitly by timestamp in expireSnapshots The previous logic relied on implicit ordering of the snapshot list. Now explicitly sorts snapshots by timestamp descending (most recent first) and uses a simpler keep-count loop: keep the first MaxSnapshotsToKeep newest snapshots plus the current snapshot unconditionally, then expire the rest that exceed the retention window. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Handle errors properly in listFilerEntries Previously all errors from ListEntries and Recv were silently swallowed. Now: treat "not found" errors as empty directory, propagate other ListEntries errors, and check for io.EOF explicitly on Recv instead of breaking on any error. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Fix overly broad HasSuffix check in orphan detection The bare strings.HasSuffix(ref, entry.Name) could match files with similar suffixes (e.g. "123.avro" matching "snap-123.avro"). Replaced with exact relPath match and a "/"-prefixed suffix check to avoid false positives. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Replace fmt.Sscanf with strconv.Atoi in extractMetadataVersion strconv.Atoi is more explicit and less fragile than fmt.Sscanf for parsing a simple integer from a trimmed string. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Recursively traverse directories for orphan file detection The orphan cleanup only listed a single directory level under data/ and metadata/, skipping IsDirectory entries. Partitioned Iceberg tables store data files in nested partition directories (e.g. data/region=us-east/file.parquet) which were never evaluated. Add walkFilerEntries helper that recursively descends into subdirectories, and use it in removeOrphans so all nested files are considered for orphan checks. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Fix manifest path drift from double time.Now() calls rewriteManifests called time.Now().UnixMilli() twice: once for the path embedded in WriteManifest and once for the filename passed to saveFilerFile. These timestamps would differ, causing the manifest's internal path reference to not match the actual saved filename. Compute the filename once and reuse it for both WriteManifest and saveFilerFile so they always reference the same path. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Add TestManifestRewritePathConsistency test Verifies that WriteManifest returns a ManifestFile whose FilePath() matches the path passed in, and that path.Base() of that path matches the filename used for saveFilerFile. This validates the single- timestamp pattern used in rewriteManifests produces consistent paths. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Make parseOperations return error on unknown operations Previously parseOperations silently dropped unknown operation names and could return an empty list. Now validates inputs against the canonical set and returns a clear error if any unknown operation is specified. Updated Execute to surface the error instead of proceeding with an empty operation list. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Use gRPC status codes instead of string matching in listFilerEntries Replace brittle strings.Contains(err.Error(), "not found") check with status.Code(err) == codes.NotFound for proper gRPC error handling. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Add stale-plan guard in commit closures for expireSnapshots and rewriteManifests Both operations plan outside the commit mutation using a snapshot ID captured from the initial metadata read. If the table head advances concurrently, the mutation would create a snapshot parented to the wrong head or remove snapshots based on a stale view. Add a guard inside each mutation closure that verifies currentMeta.CurrentSnapshot().SnapshotID still matches the planned snapshot ID. If it differs, return errStalePlan which propagates immediately (not retried, since the plan itself is invalid). Also fix rewriteManifests to derive SequenceNumber from the fresh metadata (cs.SequenceNumber) instead of the captured currentSnap. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Add compare-and-swap to updateTableMetadataXattr updateTableMetadataXattr previously re-read the entry but did not verify the metadataVersion matched what commitWithRetry had loaded. A concurrent update could be silently clobbered. Now accepts expectedVersion parameter and compares it against the stored metadataVersion before writing. Returns errMetadataVersionConflict on mismatch, which commitWithRetry treats as retryable (deletes the staged metadata file and retries with fresh state). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Export shared plugin worker helpers for use by sub-packages Export ShouldSkipDetectionByInterval, BuildExecutorActivity, and BuildDetectorActivity so the iceberg sub-package can reuse them without duplicating logic. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Refactor iceberg maintenance handler into weed/plugin/worker/iceberg package Split the 1432-line iceberg_maintenance_handler.go into focused files in a new iceberg sub-package: handler.go, config.go, detection.go, operations.go, filer_io.go, and compact.go (Phase 2 data compaction). Key changes: - Rename types to drop stutter (IcebergMaintenanceHandler → Handler, etc.) - Fix loadFileByIcebergPath to preserve nested directory paths via normalizeIcebergPath instead of path.Base which dropped subdirectories - Check SendProgress errors instead of discarding them - Add stale-plan guard to compactDataFiles commitWithRetry closure - Add "compact" operation to parseOperations canonical order - Duplicate readStringConfig/readInt64Config helpers (~20 lines) - Update worker_runtime.go to import new iceberg sub-package Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Remove iceberg_maintenance from default plugin worker job types Iceberg maintenance is not yet ready to be enabled by default. Workers can still opt in by explicitly listing iceberg_maintenance in their job types configuration. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Clamp config values to safe minimums in ParseConfig Prevents misconfiguration by enforcing minimum values using the default constants for all config fields. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Harden filer I/O: path helpers, strict CAS guard, path traversal prevention - Use path.Dir/path.Base instead of strings.SplitN in loadCurrentMetadata - Make CAS guard error on missing or unparseable metadataVersion - Add path.Clean and traversal validation in loadFileByIcebergPath Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Fix compact: single snapshot ID, oversized bin splitting, ensureFilerDir - Use single newSnapID for all manifest entries in a compaction run - Add splitOversizedBin to break bins exceeding targetSize - Make ensureFilerDir only create on NotFound, propagate other errors Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Add wildcard filters, scan limit, and context cancellation to table scanning - Use wildcard matchers (, ?) for bucket/namespace/table filters - Add limit parameter to scanTablesForMaintenance for early termination - Add ctx.Done() checks in bucket and namespace scan loops - Update filter UI descriptions and placeholders for wildcard support Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Remove dead detection interval check and validate namespace parameter - Remove ineffective ShouldSkipDetectionByInterval call with hardcoded 0 - Add namespace to required parameter validation in Execute Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Improve operations: exponential backoff, orphan matching, full file cleanup - Use exponential backoff (50ms, 100ms, 200ms, ...) in commitWithRetry - Use normalizeIcebergPath for orphan matching instead of fragile suffix check - Add collectSnapshotFiles to traverse manifest lists → manifests → data files - Delete all unreferenced files after expiring snapshots, not just manifest lists - Refactor removeOrphans to reuse collectSnapshotFiles Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * iceberg: fix ensureFilerDir to handle filer_pb.ErrNotFound sentinel filer_pb.LookupEntry converts gRPC NotFound errors to filer_pb.ErrNotFound (a plain sentinel), so status.Code() never returns codes.NotFound for that error. This caused ensureFilerDir to return an error instead of creating the directory when it didn't exist. * iceberg: clean up orphaned artifacts when compaction commit fails Track all files written during compaction (merged data files, manifest, manifest list) and delete them if the commit or any subsequent write step fails, preventing orphaned files from accumulating in the filer. * iceberg: derive tablePath from namespace/tableName when empty An empty table_path parameter would be passed to maintenance operations unchecked. Default it to path.Join(namespace, tableName) when not provided. * iceberg: make collectSnapshotFiles return error on read/parse failure Previously, errors reading manifests were logged and skipped, returning a partial reference set. This could cause incorrect delete decisions during snapshot expiration or orphan cleanup. Now the function returns an error and all callers abort when reference data is incomplete. * iceberg: include active metadata file in removeOrphans referenced set The metadataFileName returned by loadCurrentMetadata was discarded, so the active metadata file could be incorrectly treated as an orphan and deleted. Capture it and add it to the referencedFiles map. * iceberg: only retry commitWithRetry on metadata version conflicts Previously all errors from updateTableMetadataXattr triggered retries. Now only errMetadataVersionConflict causes retry; other errors (permissions, transport, malformed xattr) fail immediately. * iceberg: respect req.Limit in fakeFilerServer.ListEntries mock The mock ListEntries ignored the Limit field, so tests couldn't exercise pagination. Now it stops streaming once Limit entries have been sent. * iceberg: validate parquet schema compatibility before merging files mergeParquetFiles now compares each source file's schema against the first file's schema and aborts with a clear error if they differ, instead of blindly writing rows that could panic or produce corrupt output. * iceberg: normalize empty JobType to canonical jobType in Execute events When request.Job.JobType is empty, status events and completion messages were emitted with a blank job type. Derive a canonical value early and use it consistently in all outbound events. * iceberg: log warning on unexpected config value types in read helpers readStringConfig and readInt64Config now log a V(1) warning when they encounter an unhandled ConfigValue kind, aiding debugging of unexpected config types that silently fall back to defaults. * worker: add iceberg_maintenance to default plugin worker job types Workers using the default job types list didn't advertise the iceberg_maintenance handler despite the handler and canonical name being registered. Add it so workers pick up the handler by default. * iceberg: use defer and detached context for compaction artifact cleanup The cleanup closure used the job context which could already be canceled, and was not called on ctx.Done() early exits. Switch to a deferred cleanup with a detached context (30s timeout) so artifact deletion completes on all exit paths including context cancellation. * iceberg: use proportional jitter in commitWithRetry backoff Fixed 25ms max jitter becomes insignificant at higher retry attempts. Use 0-20% of the current backoff value instead so jitter scales with the exponential delay. * iceberg: add malformed filename cases to extractMetadataVersion test Cover edge cases like "invalid.metadata.json", "metadata.json", "", and "v.metadata.json" to ensure the function returns 0 for unparseable inputs. * iceberg: fail compaction on manifest read errors and skip delete manifests Previously, unreadable manifests were silently skipped during compaction, which could drop live files from the entry set. Now manifest read/parse errors are returned as fatal errors. Also abort compaction when delete manifests exist since the compactor does not apply deletes — carrying them through unchanged could produce incorrect results. * iceberg: use table-relative path for active metadata file in orphan scan metadataFileName was stored as a basename (e.g. "v1.metadata.json") but the orphan scanner matches against table-relative paths like "metadata/v1.metadata.json". Prefix with "metadata/" so the active metadata file is correctly recognized as referenced. * iceberg: fix MetadataBuilderFromBase location to use metadata file path The second argument to MetadataBuilderFromBase records the previous metadata file in the metadata log. Using meta.Location() (the table root) was incorrect — it must be the actual metadata file path so old metadata files can be tracked and eventually cleaned up. * iceberg: update metadataLocation and versionToken in xattr on commit updateTableMetadataXattr was only updating metadataVersion, modifiedAt, and fullMetadata but not metadataLocation or versionToken. This left catalog state inconsistent after maintenance commits — the metadataLocation still pointed to the old metadata file and the versionToken was stale. Add a newMetadataLocation parameter and regenerate the versionToken on every commit, matching the S3 Tables handler behavior. * iceberg: group manifest entries by partition spec in rewriteManifests rewriteManifests was writing all entries into a single manifest using the table's current partition spec. For spec-evolved tables where manifests reference different partition specs, this produces an invalid manifest. Group entries by the source manifest's PartitionSpecID and write one merged manifest per spec, looking up each spec from the table's PartitionSpecs list. * iceberg: remove dead code loop for non-data manifests in compaction The early abort guard at the top of compactDataFiles already ensures no delete manifests are present. The loop that copied non-data manifests into allManifests was unreachable dead code. * iceberg: use JSON encoding in partitionKey for unambiguous grouping partitionKey used fmt.Sprintf("%d=%v") joined by commas, which produces ambiguous keys when partition values contain commas or '='. Use json.Marshal for values and NUL byte as separator to eliminate collisions. * iceberg: precompute normalized reference set in removeOrphans The orphan check was O(files × refs) because it normalized each reference path inside the per-file loop. Precompute the normalized set once for O(1) lookups per candidate file. * iceberg: add artifact cleanup to rewriteManifests on commit failure rewriteManifests writes merged manifests and a manifest list to the filer before committing but did not clean them up on failure. Add the same deferred cleanup pattern used by compactDataFiles: track written artifacts and delete them if the commit does not succeed. * iceberg: pass isDeleteData=true in deleteFilerFile deleteFilerFile called DoRemove with isDeleteData=false, which only removed filer metadata and left chunk data behind on volume servers. All other data-file deletion callers in the codebase pass true. * iceberg: clean up test: remove unused snapID, simplify TestDetectWithFakeFiler Remove unused snapID variable and eliminate the unnecessary second fake filer + entry copy in TestDetectWithFakeFiler by capturing the client from the first startFakeFiler call. * fix: update TestWorkerDefaultJobTypes to expect 5 job types The test expected 4 default job types but iceberg_maintenance was added as a 5th default in a previous commit. * iceberg: document client-side CAS TOCTOU limitation in updateTableMetadataXattr Add a note explaining the race window where two workers can both pass the version check and race at UpdateEntry. The proper fix requires server-side precondition support on UpdateEntryRequest. * iceberg: remove unused sender variable in TestFullExecuteFlow * iceberg: abort compaction when multiple partition specs are present The compactor writes all entries into a single manifest using the current partition spec, which is invalid for spec-evolved tables. Detect multiple PartitionSpecIDs and skip compaction until per-spec compaction is implemented. * iceberg: validate tablePath to prevent directory traversal Sanitize the table_path parameter with path.Clean and verify it matches the expected namespace/tableName prefix to prevent path traversal attacks via crafted job parameters. * iceberg: cap retry backoff at 5s and make it context-aware The exponential backoff could grow unbounded and blocked on time.Sleep ignoring context cancellation. Cap at 5s and use a timer with select on ctx.Done so retries respect cancellation. * iceberg: write manifest list with new snapshot identity in rewriteManifests The manifest list was written with the old snapshot's ID and sequence number, but the new snapshot created afterwards used a different identity. Compute newSnapshotID and newSeqNum before writing manifests and the manifest list so all artifacts are consistent. * ec: also remove .vif file in removeEcVolumeFiles removeEcVolumeFiles cleaned up .ecx, .ecj, and shard files but not the .vif volume info file, leaving it orphaned. The .vif file lives in the data directory alongside shard files. The directory handling for index vs data files was already correct: .ecx/.ecj are removed from IdxDirectory and shard files from Directory, matching how NewEcVolume loads them. Revert "ec: also remove .vif file in removeEcVolumeFiles" This reverts commit `acc82449e1`. * iceberg: skip orphan entries with nil Attributes instead of defaulting to epoch When entry.Attributes is nil, mtime defaulted to Unix epoch (1970), making unknown-age entries appear ancient and eligible for deletion. Skip these entries instead to avoid deleting files whose age cannot be determined. * iceberg: use unique metadata filenames to prevent concurrent write clobbering Add timestamp nonce to metadata filenames (e.g. v3-1709766000.metadata.json) so concurrent writers stage to distinct files. Update extractMetadataVersion to strip the nonce suffix, and loadCurrentMetadata to read the actual filename from the metadataLocation xattr field. * iceberg: defer artifact tracking until data file builder succeeds Move the writtenArtifacts append to after NewDataFileBuilder succeeds, so a failed builder doesn't leave a stale entry for an already-deleted file in the cleanup list. * iceberg: use detached context for metadata file cleanup Use context.WithTimeout(context.Background(), 10s) when deleting staged metadata files after CAS failure, so cleanup runs even if the original request context is canceled. * test: update default job types count to include iceberg_maintenance * iceberg: use parquet.EqualNodes for structural schema comparison Replace String()-based schema comparison with parquet.EqualNodes which correctly compares types, repetition levels, and logical types. * iceberg: add nonce-suffixed filename cases to TestExtractMetadataVersion * test: assert iceberg_maintenance is present in default job types * iceberg: validate operations config early in Detect Call parseOperations in Detect so typos in the operations config fail fast before emitting proposals, matching the validation already done in Execute. * iceberg: detect chunked files in loadFileByIcebergPath Return an explicit error when a file has chunks but no inline content, rather than silently returning empty data. Data files uploaded via S3 are stored as chunks, so compaction would otherwise produce corrupt merged files. --------- Co-authored-by: Copilot <copilot@github.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>	1 week ago
Chris Lu	d89eb8267f	s3: use url.PathUnescape for X-Amz-Copy-Source header (#8545 ) * s3: use url.PathUnescape for X-Amz-Copy-Source header (#8544) The X-Amz-Copy-Source header is a URL-encoded path, not a query string. Using url.QueryUnescape incorrectly converts literal '+' characters to spaces, which can cause object key mismatches during copy operations. Switch to url.PathUnescape in CopyObjectHandler, CopyObjectPartHandler, and pathToBucketObjectAndVersion to correctly handle special characters like '!', '+', and other RFC 3986 sub-delimiters that S3 clients may percent-encode (e.g. '!' as %21). * s3: add path validation to CopyObjectPartHandler CopyObjectPartHandler was missing the validateTableBucketObjectPath checks that CopyObjectHandler has, allowing potential path traversal in the source bucket/object of copy part requests. * s3: fix case-sensitive HeadersRegexp for copy source routing The HeadersRegexp for X-Amz-Copy-Source used `%2F` which only matched uppercase hex encoding. RFC 3986 allows both `%2F` and `%2f`, so clients sending lowercase percent-encoding would bypass the copy handler and hit PutObjectHandler instead. Add (?i) flag for case-insensitive matching. Also add test coverage for the versionId branch in pathToBucketObjectAndVersion and for lowercase %2f routing.	1 week ago
Chris Lu	3f946fc0c0	mount: make metadata cache rebuilds snapshot-consistent (#8531 ) * filer: expose metadata events and list snapshots * mount: invalidate hot directory caches * mount: read hot directories directly from filer * mount: add sequenced metadata cache applier * mount: apply metadata responses through cache applier * mount: replay snapshot-consistent directory builds * mount: dedupe self metadata events * mount: factor directory build cleanup * mount: replace proto marshal dedup with composite key and ring buffer The dedup logic was doing a full deterministic proto.Marshal on every metadata event just to produce a dedup key. Replace with a cheap composite string key (TsNs\|Directory\|OldName\|NewName). Also replace the sliding-window slice (which leaked the backing array unboundedly) with a fixed-size ring buffer that reuses the same array. * filer: remove mutex and proto.Clone from request-scoped MetadataEventSink MetadataEventSink is created per-request and only accessed by the goroutine handling the gRPC call. The mutex and double proto.Clone (once in Record, once in Last) were unnecessary overhead on every filer write operation. Store the pointer directly instead. * mount: skip proto.Clone for caller-owned metadata events Add ApplyMetadataResponseOwned that takes ownership of the response without cloning. Local metadata events (mkdir, create, flush, etc.) are freshly constructed and never shared, so the clone is unnecessary. * filer: only populate MetadataEvent on successful DeleteEntry Avoid calling eventSink.Last() on error paths where the sink may contain a partial event from an intermediate child deletion during recursive deletes. * mount: avoid map allocation in collectDirectoryNotifications Replace the map with a fixed-size array and linear dedup. There are at most 3 directories to notify (old parent, new parent, new child if directory), so a 3-element array avoids the heap allocation on every metadata event. * mount: fix potential deadlock in enqueueApplyRequest Release applyStateMu before the blocking channel send. Previously, if the channel was full (cap 128), the send would block while holding the mutex, preventing Shutdown from acquiring it to set applyClosed. * mount: restore signature-based self-event filtering as fast path Re-add the signature check that was removed when content-based dedup was introduced. Checking signatures is O(1) on a small slice and avoids enqueuing and processing events that originated from this mount instance. The content-based dedup remains as a fallback. * filer: send snapshotTsNs only in first ListEntries response The snapshot timestamp is identical for every entry in a single ListEntries stream. Sending it in every response message wastes wire bandwidth for large directories. The client already reads it only from the first response. * mount: exit read-through mode after successful full directory listing MarkDirectoryRefreshed was defined but never called, so directories that entered read-through mode (hot invalidation threshold) stayed there permanently, hitting the filer on every readdir even when cold. Call it after a complete read-through listing finishes. * mount: include event shape and full paths in dedup key The previous dedup key only used Names, which could collapse distinct rename targets. Include the event shape (C/D/U/R), source directory, new parent path, and both entry names so structurally different events are never treated as duplicates. * mount: drain pending requests on shutdown in runApplyLoop After receiving the shutdown sentinel, drain any remaining requests from applyCh non-blockingly and signal each with errMetaCacheClosed so callers waiting on req.done are released. * mount: include IsDirectory in synthetic delete events metadataDeleteEvent now accepts an isDirectory parameter so the applier can distinguish directory deletes from file deletes. Rmdir passes true, Unlink passes false. * mount: fall back to synthetic event when MetadataEvent is nil In mknod and mkdir, if the filer response omits MetadataEvent (e.g. older filer without the field), synthesize an equivalent local metadata event so the cache is always updated. * mount: make Flush metadata apply best-effort after successful commit After filer_pb.CreateEntryWithResponse succeeds, the entry is persisted. Don't fail the Flush syscall if the local metadata cache apply fails — log and invalidate the directory cache instead. Also fall back to a synthetic event when MetadataEvent is nil. * mount: make Rename metadata apply best-effort The rename has already succeeded on the filer by the time we apply the local metadata event. Log failures instead of returning errors that would be dropped by the caller anyway. * mount: make saveEntry metadata apply best-effort with fallback After UpdateEntryWithResponse succeeds, treat local metadata apply as non-fatal. Log and invalidate the directory cache on failure. Also fall back to a synthetic event when MetadataEvent is nil. * filer_pb: preserve snapshotTsNs on error in ReadDirAllEntriesWithSnapshot Return the snapshot timestamp even when the first page fails, so callers receive the snapshot boundary when partial data was received. * filer: send snapshot token for empty directory listings When no entries are streamed, send a final ListEntriesResponse with only SnapshotTsNs so clients always receive the snapshot boundary. * mount: distinguish not-found vs transient errors in lookupEntry Return fuse.EIO for non-not-found filer errors instead of unconditionally returning ENOENT, so transient failures don't masquerade as missing entries. * mount: make CacheRemoteObject metadata apply best-effort The file content has already been cached successfully. Don't fail the read if the local metadata cache update fails. * mount: use consistent snapshot for readdir in direct mode Capture the SnapshotTsNs from the first loadDirectoryEntriesDirect call and store it on the DirectoryHandle. Subsequent batch loads pass this stored timestamp so all batches use the same snapshot. Also export DoSeaweedListWithSnapshot so mount can use it directly with snapshot passthrough. * filer_pb: fix test fake to send SnapshotTsNs only on first response Match the server behavior: only the first ListEntriesResponse in a page carries the snapshot timestamp, subsequent entries leave it zero. * Fix nil pointer dereference in ListEntries stream consumers Remove the empty-directory snapshot-only response from ListEntries that sent a ListEntriesResponse with Entry==nil, which crashed every raw stream consumer that assumed resp.Entry is always non-nil. Also add defensive nil checks for resp.Entry in all raw ListEntries stream consumers across: S3 listing, broker topic lookup, broker topic config, admin dashboard, topic retention, hybrid message scanner, Kafka integration, and consumer offset storage. * Add nil guards for resp.Entry in remaining ListEntries stream consumers Covers: S3 object lock check, MQ management dashboard (version/ partition/offset loops), and topic retention version loop. * Make applyLocalMetadataEvent best-effort in Link and Symlink The filer operations already succeeded; failing the syscall because the local cache apply failed is wrong. Log a warning and invalidate the parent directory cache instead. * Make applyLocalMetadataEvent best-effort in Mkdir/Rmdir/Mknod/Unlink The filer RPC already committed; don't fail the syscall when the local metadata cache apply fails. Log a warning and invalidate the parent directory cache to force a re-fetch on next access. * flushFileMetadata: add nil-fallback for metadata event and best-effort apply Synthesize a metadata event when resp.GetMetadataEvent() is nil (matching doFlush), and make the apply best-effort with cache invalidation on failure. * Prevent double-invocation of cleanupBuild in doEnsureVisited Add a cleanupDone guard so the deferred cleanup and inline error-path cleanup don't both call DeleteFolderChildren/AbortDirectoryBuild. * Fix comment: signature check is O(n) not O(1) * Prevent deferred cleanup after successful CompleteDirectoryBuild Set cleanupDone before returning from the success path so the deferred context-cancellation check cannot undo a published build. * Invalidate parent directory caches on rename metadata apply failure When applyLocalMetadataEvent fails during rename, invalidate the source and destination parent directory caches so subsequent accesses trigger a re-fetch from the filer. * Add event nil-fallback and cache invalidation to Link and Symlink Synthesize metadata events when the server doesn't return one, and invalidate parent directory caches on apply failure. * Match requested partition when scanning partition directories Parse the partition range format (NNNN-NNNN) and match against the requested partition parameter instead of using the first directory. * Preserve snapshot timestamp across empty directory listings Initialize actualSnapshotTsNs from the caller-requested value so it isn't lost when the server returns no entries. Re-add the server-side snapshot-only response for empty directories (all raw stream consumers now have nil guards for Entry). * Fix CreateEntry error wrapping to support errors.Is/errors.As Use errors.New + %w instead of %v for resp.Error so callers can unwrap the underlying error. * Fix object lock pagination: only advance on non-nil entries Move entriesReceived inside the nil check so nil entries don't cause repeated ListEntries calls with the same lastFileName. * Guard Attributes nil check before accessing Mtime in MQ management * Do not send nil-Entry response for empty directory listings The snapshot-only ListEntriesResponse (with Entry == nil) for empty directories breaks consumers that treat any received response as an entry (Java FilerClient, S3 listing). The Go client-side DoSeaweedListWithSnapshot already preserves the caller-requested snapshot via actualSnapshotTsNs initialization, so the server-side send is unnecessary. * Fix review findings: subscriber dedup, invalidation normalization, nil guards, shutdown race - Remove self-signature early-return in processEventFn so all events flow through the applier (directory-build buffering sees self-originated events that arrive after a snapshot) - Normalize NewParentPath in collectEntryInvalidations to avoid duplicate invalidations when NewParentPath is empty (same-directory update) - Guard resp.Entry.Attributes for nil in admin_server.go and topic_retention.go to prevent panics on entries without attributes - Fix enqueueApplyRequest race with shutdown by using select on both applyCh and applyDone, preventing sends after the apply loop exits - Add cleanupDone check to deferred cleanup in meta_cache_init.go for clarity alongside the existing guard in cleanupBuild - Add empty directory test case for snapshot consistency * Propagate authoritative metadata event from CacheRemoteObjectToLocalCluster and generate client-side snapshot for empty directories - Add metadata_event field to CacheRemoteObjectToLocalClusterResponse proto so the filer-emitted event is available to callers - Use WithMetadataEventSink in the server handler to capture the event from NotifyUpdateEvent and return it on the response - Update filehandle_read.go to prefer the RPC's metadata event over a locally fabricated one, falling back to metadataUpdateEvent when the server doesn't provide one (e.g., older filers) - Generate a client-side snapshot cutoff in DoSeaweedListWithSnapshot when the server sends no snapshot (empty directory), so callers like CompleteDirectoryBuild get a meaningful boundary for filtering buffered events * Skip directory notifications for dirs being built to prevent mid-build cache wipe When a metadata event is buffered during a directory build, applyMetadataSideEffects was still firing noteDirectoryUpdate for the building directory. If the directory accumulated enough updates to become "hot", markDirectoryReadThrough would call DeleteFolderChildren, wiping entries that EnsureVisited had already inserted. The build would then complete and mark the directory cached with incomplete data. Fix by using applyMetadataSideEffectsSkippingBuildingDirs for buffered events, which suppresses directory notifications for dirs currently in buildingDirs while still applying entry invalidations. * Add test for directory notification suppression during active build TestDirectoryNotificationsSuppressedDuringBuild verifies that metadata events targeting a directory under active EnsureVisited build do NOT fire onDirectoryUpdate for that directory. In production, this prevents markDirectoryReadThrough from calling DeleteFolderChildren mid-build, which would wipe entries already inserted by the listing. The test inserts an entry during a build, sends multiple metadata events for the building directory, asserts no notifications fired for it, verifies the entry survives, and confirms buffered events are replayed after CompleteDirectoryBuild. * Fix create invalidations, build guard, event shape, context, and snapshot error path - collectEntryInvalidations: invalidate FUSE kernel cache on pure create events (OldEntry==nil && NewEntry!=nil), not just updates and deletes - completeDirectoryBuildNow: only call markCachedFn when an active build existed (state != nil), preventing an unpopulated directory from being marked as cached - Add metadataCreateEvent helper that produces a create-shaped event (NewEntry only, no OldEntry) and use it in mkdir, mknod, symlink, and hardlink create fallback paths instead of metadataUpdateEvent which incorrectly set both OldEntry and NewEntry - applyMetadataResponseEnqueue: use context.Background() for the queued mutation so a cancelled caller context cannot abort the apply loop mid-write - DoSeaweedListWithSnapshot: move snapshot initialization before ListEntries call so the error path returns the preserved snapshot instead of 0 * Fix review findings: test loop, cache race, context safety, snapshot consistency - Fix build test loop starting at i=1 instead of i=0, missing new-0.txt verification - Re-check IsDirectoryCached after cache miss to avoid ENOENT race with markDirectoryReadThrough - Use context.Background() in enqueueAndWait so caller cancellation can't abort build/complete mid-way - Pass dh.snapshotTsNs in skip-batch loadDirectoryEntriesDirect for snapshot consistency - Prefer resp.MetadataEvent over fallback in Unlink event derivation - Add comment on MetadataEventSink.Record single-event assumption * Fix empty-directory snapshot clock skew and build cancellation race Empty-directory snapshot: Remove client-side time.Now() synthesis when the server returns no entries. Instead return snapshotTsNs=0, and in completeDirectoryBuildNow replay ALL buffered events when snapshot is 0. This eliminates the clock-skew bug where a client ahead of the filer would filter out legitimate post-list events. Build cancellation: Use context.Background() for BeginDirectoryBuild and CompleteDirectoryBuild calls in doEnsureVisited, so errgroup cancellation doesn't cause enqueueAndWait to return early and trigger cleanupBuild while the operation is still queued. * Add tests for empty-directory build replay and cancellation resilience TestEmptyDirectoryBuildReplaysAllBufferedEvents: verifies that when CompleteDirectoryBuild receives snapshotTsNs=0 (empty directory, no server snapshot), ALL buffered events are replayed regardless of their TsNs values — no clock-skew-sensitive filtering occurs. TestBuildCompletionSurvivesCallerCancellation: verifies that once CompleteDirectoryBuild is enqueued, a cancelled caller context does not prevent the build from completing. The apply loop runs with context.Background(), so the directory becomes cached and buffered events are replayed even when the caller gives up waiting. * Fix directory subtree cleanup, Link rollback, test robustness - applyMetadataResponseLocked: when a directory entry is deleted or moved, call DeleteFolderChildren on the old path so cached descendants don't leak as stale entries. - Link: save original HardLinkId/Counter before mutation. If CreateEntryWithResponse fails after the source was already updated, rollback the source entry to its original state via UpdateEntry. - TestBuildCompletionSurvivesCallerCancellation: replace fixed time.Sleep(50ms) with a deadline-based poll that checks IsDirectoryCached in a loop, failing only after 2s timeout. - TestReadDirAllEntriesWithSnapshotEmptyDirectory: assert that ListEntries was actually invoked on the mock client so the test exercises the RPC path. - newMetadataEvent: add early return when both oldEntry and newEntry are nil to avoid emitting events with empty Directory. --------- Co-authored-by: Copilot <copilot@github.com>	1 week ago
Chris Lu	af4c3fcb31	ec: fall back to data dir when ecx file not found in idx dir (#8541 ) * ec: fall back to data dir when ecx file not found in idx dir (#8540) When -dir.idx is configured after EC encoding, the .ecx/.ecj files remain in the data directory. NewEcVolume now falls back to the data directory when the index file is not found in dirIdx. * ec: add fallback logging and improved error message for ecx lookup * ec: preserve configured dirIdx, track actual ecx location separately The previous fallback set ev.dirIdx = dir when finding .ecx in the data directory, which corrupted IndexBaseFileName() for future writes (e.g., WriteIdxFileFromEcIndex during EC-to-volume conversion would write the .idx file to the data directory instead of the configured index directory). Introduce ecxActualDir to track where .ecx/.ecj were actually found, used only by FileName() for cleanup/destroy. IndexBaseFileName() continues to use the configured dirIdx for new file creation. * ec: check both idx and data dirs for .ecx in all cleanup and lookup paths When -dir.idx is configured after EC encoding, .ecx/.ecj files may reside in the data directory. Several code paths only checked l.IdxDirectory, causing them to miss these files: - removeEcVolumeFiles: now removes .ecx/.ecj from both directories - loadExistingVolume: ecx existence check falls back to data dir - deleteEcShardIdsForEachLocation: ecx existence check and cleanup both cover the data directory - VolumeEcShardsRebuild: ecx lookup falls back to data directory so RebuildEcxFile operates on the correct file	1 week ago
Chris Lu	540fc97e00	s3/iam: reuse one request id per request (#8538 ) * request_id: add shared request middleware * s3err: preserve request ids in responses and logs * iam: reuse request ids in XML responses * sts: reuse request ids in XML responses * request_id: drop legacy header fallback * request_id: use AWS-style request id format * iam: fix AWS-compatible XML format for ErrorResponse and field ordering - ErrorResponse uses bare <RequestId> at root level instead of <ResponseMetadata> wrapper, matching the AWS IAM error response spec - Move CommonResponse to last field in success response structs so <ResponseMetadata> serializes after result elements - Add randomness to request ID generation to avoid collisions - Add tests for XML ordering and ErrorResponse format * iam: remove duplicate error_response_test.go Test is already covered by responses_test.go. * address PR review comments - Guard against typed nil pointers in SetResponseRequestID before interface assertion (CodeRabbit) - Use regexp instead of strings.Index in test helpers for extracting request IDs (Gemini) * request_id: prevent spoofing, fix nil-error branch, thread reqID to error writers - Ensure() now always generates a server-side ID, ignoring client-sent x-amz-request-id headers to prevent request ID spoofing. Uses a private context key (contextKey{}) instead of the header string. - writeIamErrorResponse in both iamapi and embedded IAM now accepts reqID as a parameter instead of calling Ensure() internally, ensuring a single request ID per request lifecycle. - The nil-iamError branch in writeIamErrorResponse now writes a 500 Internal Server Error response instead of returning silently. - Updated tests to set request IDs via context (not headers) and added tests for spoofing prevention and context reuse. * sts: add request-id consistency assertions to ActionInBody tests * test: update admin test to expect server-generated request IDs The test previously sent a client x-amz-request-id header and expected it echoed back. Since Ensure() now ignores client headers to prevent spoofing, update the test to verify the server returns a non-empty server-generated request ID instead. * iam: add generic WithRequestID helper alongside reflection-based fallback Add WithRequestID[T] that uses generics to take the address of a value type, satisfying the pointer receiver on SetRequestId without reflection. The existing SetResponseRequestID is kept for the two call sites that operate on interface{} (from large action switches where the concrete type varies at runtime). Generics cannot replace reflection there since Go cannot infer type parameters from interface{}. * Remove reflection and generics from request ID setting Call SetRequestId directly on concrete response types in each switch branch before boxing into interface{}, eliminating the need for WithRequestID (generics) and SetResponseRequestID (reflection). * iam: return pointer responses in action dispatch * Fix IAM error handling consistency and ensure request IDs on all responses - UpdateUser/CreatePolicy error branches: use writeIamErrorResponse instead of s3err.WriteErrorResponse to preserve IAM formatting and request ID - ExecuteAction: accept reqID parameter and generate one if empty, ensuring every response carries a RequestId regardless of caller * Clean up inline policies on DeleteUser and UpdateUser rename DeleteUser: remove InlinePolicies[userName] from policy storage before removing the identity, so policies are not orphaned. UpdateUser: move InlinePolicies[userName] to InlinePolicies[newUserName] when renaming, so GetUserPolicy/DeleteUserPolicy work under the new name. Both operations persist the updated policies and return an error if the storage write fails, preventing partial state.	1 week ago
Aaron	14cd0f53ba	Places the CommonResponse struct at the end of all IAM responses. (#8537 ) * Places the CommonResponse struct at the end of all IAM responses, rather than the start. * iam: fix error response request id layout * iam: add XML ordering regression test * iam: share request id generation --------- Co-authored-by: Aaron Segal <aaron.segal@rpsolutions.com> Co-authored-by: Chris Lu <chris.lu@gmail.com>	1 week ago
Chris Lu	f9311a3422	s3api: fix static IAM policy enforcement after reload (#8532 ) * s3api: honor attached IAM policies over legacy actions * s3api: hydrate IAM policy docs during config reload * s3api: use policy-aware auth when listing buckets * credential: propagate context through filer_etc policy reads * credential: make legacy policy deletes durable * s3api: exercise managed policy runtime loader * s3api: allow static IAM users without session tokens * iam: deny unmatched attached policies under default allow * iam: load embedded policy files from filer store * s3api: require session tokens for IAM presigning * s3api: sync runtime policies into zero-config IAM * credential: respect context in policy file loads * credential: serialize legacy policy deletes * iam: align filer policy store naming * s3api: use authenticated principals for presigning * iam: deep copy policy conditions * s3api: require request creation in policy tests * filer: keep ReadInsideFiler as the context-aware API * iam: harden filer policy store writes * credential: strengthen legacy policy serialization test * credential: forward runtime policy loaders through wrapper * s3api: harden runtime policy merging * iam: require typed already-exists errors	1 week ago
Chris Lu	338be16254	fix logs	2 weeks ago
Chris Lu	1b6e96614d	s3api: cache parsed IAM policy engines for fallback auth Previously, evaluateIAMPolicies created a new PolicyEngine and re-parsed the JSON policy document for every policy on every request. This adds a shared iamPolicyEngine field that caches compiled policies, kept in sync by PutPolicy, DeletePolicy, and bulk config reload paths. - PutPolicy deletes the old cache entry before setting the new one, so a parse failure on update does not leave a stale allow. - Log warnings when policy compilation fails instead of silently discarding errors. - Add test for valid-to-invalid policy update regression.	2 weeks ago
SrikanthBhandary	4eb45ecc5e	s3api: add IAM policy fallback authorization tests (#8518 ) * s3api: add IAM policy fallback auth with tests * s3api: use policy engine for IAM fallback evaluation	2 weeks ago
Chris Lu	1f3df6e9ef	admin: remove Alpha badge and unused Metrics/Logs menu items (#8525 ) * admin: remove Alpha badge and unused Metrics/Logs menu items * Update layout_templ.go	2 weeks ago
Chris Lu	230ae9c24e	no need to set default scripts now	2 weeks ago
Chris Lu	b3f7472fd3	4.15	2 weeks ago
Chris Lu	b3620c7e14	admin: auto migrating master maintenance scripts to admin_script plugin config (#8509 ) * admin: seed admin_script plugin config from master maintenance scripts When the admin server starts, fetch the maintenance scripts configuration from the master via GetMasterConfiguration. If the admin_script plugin worker does not already have a saved config, use the master's scripts as the default value. This enables seamless migration from master.toml [master.maintenance] to the admin script plugin worker. Changes: - Add maintenance_scripts and maintenance_sleep_minutes fields to GetMasterConfigurationResponse in master.proto - Populate the new fields from viper config in master_grpc_server.go - On admin server startup, fetch the master config and seed the admin_script plugin config if no config exists yet - Strip lock/unlock commands from the master scripts since the admin script worker handles locking automatically Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: address review comments on admin_script seeding - Replace TOCTOU race (separate Load+Save) with atomic SaveJobTypeConfigIfNotExists on ConfigStore and Plugin - Replace ineffective polling loop with single GetMaster call using 30s context timeout, since GetMaster respects context cancellation - Add unit tests for SaveJobTypeConfigIfNotExists (in-memory + on-disk) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: apply maintenance script defaults in gRPC handler The gRPC handler for GetMasterConfiguration read maintenance scripts from viper without calling SetDefault, relying on startAdminScripts having run first. If the admin server calls GetMasterConfiguration before startAdminScripts sets the defaults, viper returns empty strings and the seeding is silently skipped. Apply SetDefault in the gRPC handler itself so it is self-contained. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * Revert "fix: apply maintenance script defaults in gRPC handler" This reverts commit `068a506330`. * fix: use atomic save in ensureJobTypeConfigFromDescriptor ensureJobTypeConfigFromDescriptor used a separate Load + Save, racing with seedAdminScriptFromMaster. If the descriptor defaults (empty script) were saved first, SaveJobTypeConfigIfNotExists in the seeding goroutine would see an existing config and skip, losing the master's maintenance scripts. Switch to SaveJobTypeConfigIfNotExists so both paths are atomic. Whichever wins, the other is a safe no-op. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: fetch master scripts inline during config bootstrap, not in goroutine Replace the seedAdminScriptFromMaster goroutine with a ConfigDefaultsProvider callback. When the plugin bootstraps admin_script defaults from the worker descriptor, it calls the provider which fetches maintenance scripts from the master synchronously. This eliminates the race between the seeding goroutine and the descriptor-based config bootstrap. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * skip commented lock unlock Co-Authored-By: Copilot <223556219+Copilot@users.noreply.github.com> * reduce grpc calls --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>	2 weeks ago
Chris Lu	7799804200	4.14 Co-Authored-By: Copilot <223556219+Copilot@users.noreply.github.com>	2 weeks ago
Chris Lu	c19f88eef1	fix: resolve ServerAddress to NodeId in maintenance task sync (#8508 ) * fix: maintenance task topology lookup, retry, and stale task cleanup 1. Strip gRPC port from ServerAddress in SyncTask using ToHttpAddress() so task targets match topology disk keys (NodeId format). 2. Skip capacity check when topology has no disks yet (startup race where tasks are loaded from persistence before first topology update). 3. Don't retry permanent errors like "volume not found" - these will never succeed on retry. 4. Cancel all pending tasks for each task type before re-detection, ensuring stale proposals from previous cycles are cleaned up. This prevents stale tasks from blocking new detection and from repeatedly failing. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * logs Co-Authored-By: Copilot <223556219+Copilot@users.noreply.github.com> * less lock scope Co-Authored-By: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>	2 weeks ago
Fábio Henrique Araújo	88e8342e44	style: Reseted padding to container-fluid div in layout template (#8505 ) * style: Reseted padding to container-fluid div in layout template * address comment Co-Authored-By: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Chris Lu <chris.lu@gmail.com> Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>	2 weeks ago
Chris Lu	df5e8210df	Implement IAM managed policy operations (#8507 ) * feat: Implement IAM managed policy operations (GetPolicy, ListPolicies, DeletePolicy, AttachUserPolicy, DetachUserPolicy) - Add response type aliases in iamapi_response.go for managed policy operations - Implement 6 handler methods in iamapi_management_handlers.go: - GetPolicy: Lookup managed policy by ARN - DeletePolicy: Remove managed policy - ListPolicies: List all managed policies - AttachUserPolicy: Attach managed policy to user, aggregating inline + managed actions - DetachUserPolicy: Detach managed policy from user - ListAttachedUserPolicies: List user's attached managed policies - Add computeAllActionsForUser() to aggregate actions from both inline and managed policies - Wire 6 new DoActions switch cases for policy operations - Add comprehensive tests for all new handlers - Fixes #8506 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: address PR review feedback for IAM managed policy operations - Add parsePolicyArn() helper with proper ARN prefix validation, replacing fragile strings.Split parsing in GetPolicy, DeletePolicy, AttachUserPolicy, and DetachUserPolicy - DeletePolicy now detaches the policy from all users and recomputes their aggregated actions, preventing stale permissions after deletion - Set changed=true for DeletePolicy DoActions case so identity updates persist - Make PolicyId consistent: CreatePolicy now uses Hash(&policyName) matching GetPolicy and ListPolicies - Remove redundant nil map checks (Go handles nil map lookups safely) - DRY up action deduplication in computeAllActionsForUser with addUniqueActions closure - Add tests for invalid/empty ARN rejection and DeletePolicy identity cleanup Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat: add integration tests for managed policy lifecycle (#8506) Add two integration tests covering the user-reported use case where managed policy operations returned 500 errors: - TestS3IAMManagedPolicyLifecycle: end-to-end workflow matching the issue report — CreatePolicy, ListPolicies, GetPolicy, AttachUserPolicy, ListAttachedUserPolicies, idempotent re-attach, DeletePolicy while attached (expects DeleteConflict), DetachUserPolicy, DeletePolicy, and verification that deleted policy is gone - TestS3IAMManagedPolicyErrorCases: covers error paths — nonexistent policy/user for GetPolicy, DeletePolicy, AttachUserPolicy, DetachUserPolicy, and ListAttachedUserPolicies Also fixes DeletePolicy to reject deletion when policy is still attached to a user (AWS-compatible DeleteConflictException), and adds the 409 status code mapping for DeleteConflictException in the error response handler. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: nil map panic in CreatePolicy, add PolicyId test assertions - Initialize policies.Policies map in CreatePolicy if nil (prevents panic when no policies exist yet); also handle filer_pb.ErrNotFound like other callers - Add PolicyId assertions in TestGetPolicy and TestListPolicies to lock in the consistent Hash(&policyName) behavior - Remove redundant time.Sleep calls from new integration tests (startMiniCluster already blocks on waitForS3Ready) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: PutUserPolicy and DeleteUserPolicy now preserve managed policy actions PutUserPolicy and DeleteUserPolicy were calling computeAggregatedActionsForUser (inline-only), overwriting ident.Actions and dropping managed policy actions. Both now call computeAllActionsForUser which unions inline + managed actions. Add TestManagedPolicyActionsPreservedAcrossInlineMutations regression test: attaches a managed policy, adds an inline policy (verifies both actions present), deletes the inline policy, then asserts managed policy actions still persist. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: PutUserPolicy verifies user exists before persisting inline policy Previously the inline policy was written to storage before checking if the target user exists in s3cfg.Identities, leaving orphaned policy data when the user was absent. Now validates the user first, returning NoSuchEntityException immediately if not found. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: prevent stale/lost actions on computeAllActionsForUser failure - PutUserPolicy: on recomputation failure, preserve existing ident.Actions instead of falling back to only the current inline policy's actions - DeleteUserPolicy: on recomputation failure, preserve existing ident.Actions instead of assigning nil (which wiped all permissions) - AttachUserPolicy: roll back ident.PolicyNames and return error if action recomputation fails, keeping identity consistent - DetachUserPolicy: roll back ident.PolicyNames and return error if GetPolicies or action recomputation fails - Add doc comment on newTestIamApiServer noting it only sets s3ApiConfig Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>	2 weeks ago
Chris Lu	10a30a83e1	s3api: add GetObjectAttributes API support (#8504 ) * s3api: add error code and header constants for GetObjectAttributes Add ErrInvalidAttributeName error code and header constants (X-Amz-Object-Attributes, X-Amz-Max-Parts, X-Amz-Part-Number-Marker, X-Amz-Delete-Marker) needed by the S3 GetObjectAttributes API. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: implement GetObjectAttributes handler Add GetObjectAttributesHandler that returns selected object metadata (ETag, Checksum, StorageClass, ObjectSize, ObjectParts) without returning the object body. Follows the same versioning and conditional header patterns as HeadObjectHandler. The handler parses the X-Amz-Object-Attributes header to determine which attributes to include in the XML response, and supports ObjectParts pagination via X-Amz-Max-Parts and X-Amz-Part-Number-Marker. Ref: https://docs.aws.amazon.com/AmazonS3/latest/API/API_GetObjectAttributes.html Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: register GetObjectAttributes route Register the GET /{object}?attributes route for the GetObjectAttributes API, placed before other object query routes to ensure proper matching. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: add integration tests for GetObjectAttributes Test coverage: - Basic: simple object with all attribute types - MultipartObject: multipart upload with parts pagination - SelectiveAttributes: requesting only specific attributes - InvalidAttribute: server rejects invalid attribute names - NonExistentObject: returns NoSuchKey for missing objects Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: add versioned object test for GetObjectAttributes Test puts two versions of the same object and verifies that: - GetObjectAttributes returns the latest version by default - GetObjectAttributes with versionId returns the specific version - ObjectSize and VersionId are correct for each version Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: fix combined conditional header evaluation per RFC 7232 Per RFC 7232: - Section 3.4: If-Unmodified-Since MUST be ignored when If-Match is present (If-Match is the more accurate replacement) - Section 3.3: If-Modified-Since MUST be ignored when If-None-Match is present (If-None-Match is the more accurate replacement) Previously, all four conditional headers were evaluated independently. This caused incorrect 412 responses when If-Match succeeded but If-Unmodified-Since failed (should return 200 per AWS S3 behavior). Fix applied to both validateConditionalHeadersForReads (GET/HEAD) and validateConditionalHeaders (PUT) paths. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: add conditional header combination tests for GetObjectAttributes Test the RFC 7232 combined conditional header semantics: - If-Match=true + If-Unmodified-Since=false => 200 (If-Unmodified-Since ignored) - If-None-Match=false + If-Modified-Since=true => 304 (If-Modified-Since ignored) - If-None-Match=true + If-Modified-Since=false => 200 (If-Modified-Since ignored) - If-Match=true + If-Unmodified-Since=true => 200 - If-Match=false => 412 regardless Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: document Checksum attribute as not yet populated Checksum is accepted in validation (so clients requesting it don't get a 400 error, matching AWS behavior for objects without checksums) but SeaweedFS does not yet store S3 checksums. Add a comment explaining this and noting where to populate it when checksum storage is added. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: add s3:GetObjectAttributes IAM action for ?attributes query Previously, GET /{object}?attributes resolved to s3:GetObject via the fallback path since resolveFromQueryParameters had no case for the "attributes" query parameter. Add S3_ACTION_GET_OBJECT_ATTRIBUTES constant ("s3:GetObjectAttributes") and a branch in resolveFromQueryParameters to return it for GET requests with the "attributes" query parameter, so IAM policies can distinguish GetObjectAttributes from GetObject. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: evaluate conditional headers after version resolution Move conditional header evaluation (If-Match, If-None-Match, etc.) to after the version resolution step in GetObjectAttributesHandler. This ensures that when a specific versionId is requested, conditions are checked against the correct version entry rather than always against the latest version. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: use bounded HTTP client in GetObjectAttributes tests Replace http.DefaultClient with a timeout-aware http.Client (10s) in the signedGetObjectAttributes helper and testGetObjectAttributesInvalid to prevent tests from hanging indefinitely. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: check attributes query before versionId in action resolver Move the GetObjectAttributes action check before the versionId check in resolveFromQueryParameters. This fixes GET /bucket/key?attributes&versionId=xyz being incorrectly classified as s3:GetObjectVersion instead of s3:GetObjectAttributes. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: add tests for versioned conditional headers and action resolver Add integration test that verifies conditional headers (If-Match, If-None-Match) are evaluated against the requested version entry, not the latest version. This covers the fix in `55c409dec`. Add unit test for ResolveS3Action verifying that the attributes query parameter takes precedence over versionId, so GET ?attributes&versionId resolves to s3:GetObjectAttributes. This covers the fix in `b92c61c95`. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: guard negative chunk indices and rename PartsCount field Add bounds checks for b.StartChunk >= 0 and b.EndChunk >= 0 in buildObjectAttributesParts to prevent panics from corrupted metadata with negative index values. Rename ObjectAttributesParts.PartsCount to TotalPartsCount to match the AWS SDK v2 Go field naming convention, while preserving the XML element name "PartsCount" via the struct tag. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * s3api: reject malformed max-parts and part-number-marker headers Return ErrInvalidMaxParts and ErrInvalidPartNumberMarker when the X-Amz-Max-Parts or X-Amz-Part-Number-Marker headers contain non-integer or negative values, matching ListObjectPartsHandler behavior. Previously these were silently ignored with defaults. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>	2 weeks ago
Racci	9e26d6f5dd	fix: port in SNI address when using domainName instead of IP for master (#8500 )	2 weeks ago
Copilot	70ed9c2a55	Update plugin_templ.go Co-Authored-By: Copilot <223556219+Copilot@users.noreply.github.com>	2 weeks ago

1 2 3 4 5 ...

8419 Commits (737116e83c3fb2362b0bfb3c242ed95a00a511ea)