Commit Graph

9 Commits

Author SHA1 Message Date
Joseph Doherty
4de691c9c5 perf: add FileStore buffered writes, O(1) state tracking, and eliminate redundant per-publish work
Implement Go-parity background flush loop (coalesce 16KB/8ms) in MsgBlock/FileStore,
replace O(n) GetStateAsync with incremental counters, skip PruneExpired/LoadAsync/
PrunePerSubject when not needed, and bypass RAFT for single-replica streams. Fix counter
tracking bugs in RemoveMsg/EraseMsg/TTL expiry and ObjectDisposedException races in
flush loop disposal. FileStore optimizations verified with 3112/3112 JetStream tests
passing; async publish benchmark remains at ~174 msg/s due to E2E protocol path bottleneck.
2026-03-13 03:11:11 -04:00
Joseph Doherty
cbe41d0efb feat: add SequenceSet for sparse deletion tracking with secure erase (Gap 1.7)
Replace HashSet<ulong> _deleted in MsgBlock with SequenceSet — a sorted-range
list that compresses contiguous deletions into (Start, End) intervals. Adds
O(log n) Contains/Add via binary search on range count, matching Go's avl.SequenceSet
semantics with a simpler implementation.

- Add SequenceSet.cs: sorted-range compressed set with Add/Remove/Contains/Count/Clear
  and IEnumerable<ulong> in ascending order. Binary search for all O(log n) ops.
- Replace HashSet<ulong> _deleted and _skipSequences in MsgBlock with SequenceSet.
- Add secureErase parameter (default false) to MsgBlock.Delete(): when true, payload
  bytes are overwritten with RandomNumberGenerator.Fill() before the delete record is
  written, making original content unrecoverable on disk.
- Update FileStore.DeleteInBlock() to propagate secureErase flag.
- Update FileStore.EraseMsg() to use secureErase: true via block layer instead of
  delegating to RemoveMsg().
- Add SequenceSetTests.cs: 25 tests covering Add, Remove, Contains, Count, range
  compression, gap filling, bridge merges, enumeration, boundary values, round-trip.
- Add FileStoreTombstoneTrackingTests.cs: 12 tests covering SequenceSet tracking in
  MsgBlock, tombstone persistence through RebuildIndex recovery, secure erase
  payload overwrite verification, and FileStore.EraseMsg integration.

Go reference: filestore.go:5267 (removeMsg), filestore.go:5890 (eraseMsg),
              avl/seqset.go (SequenceSet).
2026-02-25 08:02:44 -05:00
Joseph Doherty
5beeb1b3f6 feat: add checksum validation on MsgBlock read path (Gap 1.5)
Add _lastChecksum field and LastChecksum property to MsgBlock tracking
the XxHash64 checksum of the last written record (Go: msgBlock.lchk,
filestore.go:2204). Capture the checksum from the encoded record trailer
on every Write/WriteAt/WriteSkip call. Read-path validation happens
naturally through the existing MessageRecord.Decode checksum check.
2026-02-25 07:50:03 -05:00
Joseph Doherty
7eb06c8ac5 feat: add FileStore tombstone, TTL & consumer state persistence (Task 2)
Port Go filestore tombstone/deletion tests, consumer state encode/decode,
consumer file store persistence, and message TTL enforcement. Adds
ConsumerStateCodec and ConsumerFileStore implementations.

17 new tests ported from filestore_test.go.
2026-02-24 20:17:35 -05:00
Joseph Doherty
a245bd75a7 feat(storage): port FileStore Go tests and add sync methods (Go parity)
Add 67 Go-parity tests from filestore_test.go covering:
- SkipMsg/SkipMsgs sequence reservation
- RemoveMsg/EraseMsg soft-delete
- LoadMsg/LoadLastMsg/LoadNextMsg message retrieval
- AllLastSeqs/MultiLastSeqs per-subject last sequences
- SubjectForSeq reverse lookup
- NumPending with filters and last-per-subject mode
- Recovery watermark preservation after purge
- FastState NumDeleted/LastTime correctness
- PurgeEx with empty subject + keep parameter
- Compact _first watermark tracking
- Multi-block operations and state verification

Implements missing IStreamStore sync methods on FileStore:
RemoveMsg, EraseMsg, SkipMsg, SkipMsgs, LoadMsg, LoadLastMsg,
LoadNextMsg, AllLastSeqs, MultiLastSeqs, SubjectForSeq, NumPending.

Adds MsgBlock.WriteSkip() for tombstone sequence reservation.
Adds IDisposable to FileStore for synchronous test disposal.
2026-02-24 14:43:06 -05:00
Joseph Doherty
d0068b121f feat(storage): add write cache and TTL scheduling (Go parity)
Add MsgBlock write cache (mirrors Go's msgBlock.cache) to serve reads
for recently-written records without disk I/O; cleared on block seal via
RotateBlock. Add HashWheel-based TTL expiry in FileStore (ExpireFromWheel /
RegisterTtl), replacing the O(n) linear scan on every append with an
O(expired) wheel scan. Implement StoreMsg sync method with per-message TTL
override support. Add 10 tests covering cache hits/eviction, wheel expiry,
retention, StoreMsg seq/ts, per-msg TTL, and recovery re-registration.
2026-02-24 13:54:37 -05:00
Joseph Doherty
b0b64292b3 feat(storage): add tombstone tracking and purge operations (Go parity)
Implement PurgeEx, Compact, Truncate, FilteredState, SubjectsState,
SubjectsTotals, State, FastState, GetSeqFromTime on FileStore. Add
MsgBlock.IsDeleted, DeletedSequences, EnumerateNonDeleted. Includes
wildcard subject support via SubjectMatch for all filtered operations.
2026-02-24 13:45:17 -05:00
Joseph Doherty
2816e8f048 feat(storage): rewrite FileStore to use block-based MsgBlock storage
Replace JSONL persistence with real MsgBlock-based block files (.blk).
FileStore now acts as a block manager that creates, seals, and rotates
MsgBlocks while maintaining an in-memory cache for fast reads/queries.

Key changes:
- AppendAsync writes transformed payloads to MsgBlock via WriteAt
- Block rotation occurs when active block reaches size limit
- Recovery scans .blk files and rebuilds in-memory state from records
- Legacy JSONL migration: existing messages.jsonl data is automatically
  converted to block files on first open, then JSONL is deleted
- PurgeAsync disposes and deletes all block files
- RewriteBlocks rebuilds blocks from cache (used by trim/restore)
- InvalidDataException propagates during recovery (wrong encryption key)

MsgBlock.WriteAt added to support explicit sequence numbers and timestamps,
needed when rewriting blocks with non-contiguous sequences (after removes).

Tests updated:
- New FileStoreBlockTests.cs with 9 tests for block-specific behavior
- JetStreamFileStoreCompressionEncryptionParityTests updated to read
  FSV1 magic from .blk files instead of messages.jsonl
- JetStreamFileStoreDurabilityParityTests updated to verify .blk files
  instead of index.manifest.json

All 3,562 tests pass (3,535 passed + 27 skipped, 0 failures).
2026-02-24 12:39:32 -05:00
Joseph Doherty
09252b8c79 feat(storage): add MsgBlock block-based message storage unit
MsgBlock is the unit of storage in the file store — a single append-only
block file containing sequentially written binary message records. Blocks
are sealed (read-only) when they reach a configurable byte-size limit.

Key features:
- Write: appends MessageRecord-encoded messages with auto-incrementing
  sequence numbers and configurable first sequence offset
- Read: positional I/O via RandomAccess.Read for concurrent reader safety
- Delete: soft-delete with on-disk persistence (re-encodes flags byte +
  checksum so deletions survive recovery)
- Recovery: rebuilds in-memory index by scanning block file using
  MessageRecord.MeasureRecord for record boundary detection
- Thread safety: ReaderWriterLockSlim allows concurrent reads during writes

Also adds MessageRecord.MeasureRecord() — computes a record's byte length
by parsing varint field headers without full decode, needed for sequential
record scanning during block recovery.

Reference: golang/nats-server/server/filestore.go:217-267 (msgBlock struct)

12 tests covering write, read, delete, seal, recovery, concurrency,
and custom sequence offsets.
2026-02-24 12:21:33 -05:00