* mount: improve read throughput with parallel chunk fetching This addresses issue #7504 where a single weed mount FUSE instance does not fully utilize node network bandwidth when reading large files. Changes: - Add -concurrentReaders mount option (default: 16) to control the maximum number of parallel chunk fetches during read operations - Implement parallel section reading in ChunkGroup.ReadDataAt() using errgroup for better throughput when reading across multiple sections - Enhance ReaderCache with MaybeCacheMany() to prefetch multiple chunks ahead in parallel during sequential reads (now prefetches 4 chunks) - Increase ReaderCache limit dynamically based on concurrentReaders to support higher read parallelism The bottleneck was that chunks were being read sequentially even when they reside on different volume servers. By introducing parallel chunk fetching, a single mount instance can now better saturate available network bandwidth. Fixes: #7504 * fmt * Address review comments: make prefetch configurable, improve error handling Changes: 1. Add DefaultPrefetchCount constant (4) to reader_at.go 2. Add GetPrefetchCount() method to ChunkGroup that derives prefetch count from concurrentReaders (1/4 ratio, min 1, max 8) 3. Pass prefetch count through NewChunkReaderAtFromClient 4. Fix error handling in readDataAtParallel to prioritize errgroup error 5. Update all callers to use DefaultPrefetchCount constant For mount operations, prefetch scales with -concurrentReaders: - concurrentReaders=16 (default) -> prefetch=4 - concurrentReaders=32 -> prefetch=8 (capped) - concurrentReaders=4 -> prefetch=1 For non-mount paths (WebDAV, query engine, MQ), uses DefaultPrefetchCount. * fmt * Refactor: use variadic parameter instead of new function name Use NewChunkGroup with optional concurrentReaders parameter instead of creating a separate NewChunkGroupWithConcurrency function. This maintains backward compatibility - existing callers without the parameter get the default of 16 concurrent readers. * Use explicit concurrentReaders parameter instead of variadic * Refactor: use MaybeCache with count parameter instead of new MaybeCacheMany function * Address nitpick review comments - Add upper bound (128) on concurrentReaders to prevent excessive goroutine fan-out - Cap readerCacheLimit at 256 accordingly - Fix SetChunks: use Lock() instead of RLock() since we are writing to group.sections
175 lines
4.5 KiB
Go
175 lines
4.5 KiB
Go
package mount
|
|
|
|
import (
|
|
"os"
|
|
"sync"
|
|
|
|
"github.com/seaweedfs/seaweedfs/weed/filer"
|
|
"github.com/seaweedfs/seaweedfs/weed/glog"
|
|
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
|
|
"github.com/seaweedfs/seaweedfs/weed/util"
|
|
)
|
|
|
|
type FileHandleId uint64
|
|
|
|
var IsDebugFileReadWrite = false
|
|
|
|
type FileHandle struct {
|
|
fh FileHandleId
|
|
counter int64
|
|
entry *LockedEntry
|
|
entryLock sync.RWMutex
|
|
entryChunkGroup *filer.ChunkGroup
|
|
inode uint64
|
|
wfs *WFS
|
|
|
|
// cache file has been written to
|
|
dirtyMetadata bool
|
|
dirtyPages *PageWriter
|
|
reader *filer.ChunkReadAt
|
|
contentType string
|
|
|
|
isDeleted bool
|
|
|
|
// RDMA chunk offset cache for performance optimization
|
|
chunkOffsetCache []int64
|
|
chunkCacheValid bool
|
|
chunkCacheLock sync.RWMutex
|
|
|
|
// for debugging
|
|
mirrorFile *os.File
|
|
}
|
|
|
|
func newFileHandle(wfs *WFS, handleId FileHandleId, inode uint64, entry *filer_pb.Entry) *FileHandle {
|
|
fh := &FileHandle{
|
|
fh: handleId,
|
|
counter: 1,
|
|
inode: inode,
|
|
wfs: wfs,
|
|
}
|
|
// dirtyPages: newContinuousDirtyPages(file, writeOnly),
|
|
fh.dirtyPages = newPageWriter(fh, wfs.option.ChunkSizeLimit)
|
|
fh.entry = &LockedEntry{
|
|
Entry: entry,
|
|
}
|
|
if entry != nil {
|
|
fh.SetEntry(entry)
|
|
}
|
|
|
|
if IsDebugFileReadWrite {
|
|
var err error
|
|
fh.mirrorFile, err = os.OpenFile("/tmp/sw/"+entry.Name, os.O_RDWR|os.O_CREATE, 0600)
|
|
if err != nil {
|
|
println("failed to create mirror:", err.Error())
|
|
}
|
|
}
|
|
|
|
return fh
|
|
}
|
|
|
|
func (fh *FileHandle) FullPath() util.FullPath {
|
|
fp, _ := fh.wfs.inodeToPath.GetPath(fh.inode)
|
|
return fp
|
|
}
|
|
|
|
func (fh *FileHandle) GetEntry() *LockedEntry {
|
|
return fh.entry
|
|
}
|
|
|
|
func (fh *FileHandle) SetEntry(entry *filer_pb.Entry) {
|
|
if entry != nil {
|
|
fileSize := filer.FileSize(entry)
|
|
entry.Attributes.FileSize = fileSize
|
|
var resolveManifestErr error
|
|
fh.entryChunkGroup, resolveManifestErr = filer.NewChunkGroup(fh.wfs.LookupFn(), fh.wfs.chunkCache, entry.Chunks, fh.wfs.option.ConcurrentReaders)
|
|
if resolveManifestErr != nil {
|
|
glog.Warningf("failed to resolve manifest chunks in %+v", entry)
|
|
}
|
|
} else {
|
|
glog.Fatalf("setting file handle entry to nil")
|
|
}
|
|
fh.entry.SetEntry(entry)
|
|
|
|
// Invalidate chunk offset cache since chunks may have changed
|
|
fh.invalidateChunkCache()
|
|
}
|
|
|
|
func (fh *FileHandle) UpdateEntry(fn func(entry *filer_pb.Entry)) *filer_pb.Entry {
|
|
result := fh.entry.UpdateEntry(fn)
|
|
|
|
// Invalidate chunk offset cache since entry may have been modified
|
|
fh.invalidateChunkCache()
|
|
|
|
return result
|
|
}
|
|
|
|
func (fh *FileHandle) AddChunks(chunks []*filer_pb.FileChunk) {
|
|
fh.entry.AppendChunks(chunks)
|
|
|
|
// Invalidate chunk offset cache since new chunks were added
|
|
fh.invalidateChunkCache()
|
|
}
|
|
|
|
func (fh *FileHandle) ReleaseHandle() {
|
|
|
|
fhActiveLock := fh.wfs.fhLockTable.AcquireLock("ReleaseHandle", fh.fh, util.ExclusiveLock)
|
|
defer fh.wfs.fhLockTable.ReleaseLock(fh.fh, fhActiveLock)
|
|
|
|
fh.dirtyPages.Destroy()
|
|
if IsDebugFileReadWrite {
|
|
fh.mirrorFile.Close()
|
|
}
|
|
}
|
|
|
|
func lessThan(a, b *filer_pb.FileChunk) bool {
|
|
if a.ModifiedTsNs == b.ModifiedTsNs {
|
|
return a.Fid.FileKey < b.Fid.FileKey
|
|
}
|
|
return a.ModifiedTsNs < b.ModifiedTsNs
|
|
}
|
|
|
|
// getCumulativeOffsets returns cached cumulative offsets for chunks, computing them if necessary
|
|
func (fh *FileHandle) getCumulativeOffsets(chunks []*filer_pb.FileChunk) []int64 {
|
|
fh.chunkCacheLock.RLock()
|
|
if fh.chunkCacheValid && len(fh.chunkOffsetCache) == len(chunks)+1 {
|
|
// Cache is valid and matches current chunk count
|
|
result := make([]int64, len(fh.chunkOffsetCache))
|
|
copy(result, fh.chunkOffsetCache)
|
|
fh.chunkCacheLock.RUnlock()
|
|
return result
|
|
}
|
|
fh.chunkCacheLock.RUnlock()
|
|
|
|
// Need to compute/recompute cache
|
|
fh.chunkCacheLock.Lock()
|
|
defer fh.chunkCacheLock.Unlock()
|
|
|
|
// Double-check in case another goroutine computed it while we waited for the lock
|
|
if fh.chunkCacheValid && len(fh.chunkOffsetCache) == len(chunks)+1 {
|
|
result := make([]int64, len(fh.chunkOffsetCache))
|
|
copy(result, fh.chunkOffsetCache)
|
|
return result
|
|
}
|
|
|
|
// Compute cumulative offsets
|
|
cumulativeOffsets := make([]int64, len(chunks)+1)
|
|
for i, chunk := range chunks {
|
|
cumulativeOffsets[i+1] = cumulativeOffsets[i] + int64(chunk.Size)
|
|
}
|
|
|
|
// Cache the result
|
|
fh.chunkOffsetCache = make([]int64, len(cumulativeOffsets))
|
|
copy(fh.chunkOffsetCache, cumulativeOffsets)
|
|
fh.chunkCacheValid = true
|
|
|
|
return cumulativeOffsets
|
|
}
|
|
|
|
// invalidateChunkCache invalidates the chunk offset cache when chunks are modified
|
|
func (fh *FileHandle) invalidateChunkCache() {
|
|
fh.chunkCacheLock.Lock()
|
|
fh.chunkCacheValid = false
|
|
fh.chunkOffsetCache = nil
|
|
fh.chunkCacheLock.Unlock()
|
|
}
|