Files
seaweedfs/weed/command/volume.go
T
Chris Lu 9658f309d2 EC bitrot detection: per-shard checksum sidecars (#9761)
* ec: add EC bitrot checksum protobuf

EcBitrotProtection/EcShardChecksums/ChecksumAlgorithm sidecar messages,
copy_ecsum_file and unsafe_ignore_sidecar fields, and a CHECKSUM scrub mode.

* ec: bitrot checksum sidecar format, validation, and per-volume load

Per-shard CRC32C block checksums in an optional <base>.ecsum sidecar with a
self-integrity header; validation, rolling builder, backfill primitive, and
EcVolume load on mount + removal on destroy.

* ec: capture per-shard checksums at encode; verify-and-exclude on rebuild

WriteEcFilesWithContext returns the protection computed inline during encoding.
generateMissingEcFiles verifies present inputs against the sidecar, excludes
corrupt ones, regenerates in place, and re-verifies; fail-closed unless
unsafe_ignore_sidecar, removing all generated outputs on failure.

* ec: read-only checksum scrub with Reed-Solomon arbiter

ChecksumScrub verifies each local shard against the sidecar and reconstructs
flagged shards from the clean shards so stale-sidecar false positives are not
reported. Wired to the gRPC CHECKSUM mode and ec.scrub -mode checksum.

* ec: server-side bitrot sidecar write, copy, cleanup, and opportunistic backfill

Write .ecsum at fresh encode; propagate it with copy_ecsum_file (tolerant);
remove it on full delete and decode; rebuild honors unsafe_ignore_sidecar and
opportunistically backfills a sidecar when all shards are reachable.

* ec: volume server bitrot config flags

-ec.bitrotChecksum (default on) and -ec.bitrotBlockSizeMB (default 16).

* fix(ec_bitrot): bound -ec.bitrotBlockSizeMB before the int64 multiply

Validate the MiB value is in [1, 1024] before multiplying by 1 MiB, so a huge
flag value cannot overflow int64 and slip past the power-of-two check, and a
block size cannot collapse a sidecar to a few oversized blocks.

* fix(ec_bitrot): distribute the .ecsum sidecar from the worker encode path

The worker EC encode wrote the generation-0 sidecar locally but never added it
to shardFiles, so DistributeEcShards never shipped it and the distributed
holders came up unprotected. Append it to shardFiles and map the ecsum shard
type to its extension in the sender so it travels with the shards.

* fix(ec_bitrot): remove orphaned sidecars when the generation is gone

Gate sidecar removal on existingShardCount==0 alone rather than also requiring a
stray .ecx. A sidecar whose shards have all been deleted is orphaned and must be
removed even when no .ecx remains, or it leaks. .ecx/.ecj/.vif removal stays
gated on hasEcxFile as before.

* fix(ec_bitrot): do not fold checksum blocks scanned into TotalFiles

ChecksumScrub's first return is blocks scanned, not files. Discard it so the
scrub response's TotalFiles (a needle/file count) is not inflated by the block
count for CHECKSUM mode.

* test(ec_bitrot): clean up generated .ecsum sidecars in removeGeneratedFiles

* fix(ec_bitrot): reject an oversized sidecar payload before the uint32 cast

The header stores payload_len as a uint32; bound the payload before the
conversion so a pathological manifest cannot truncate the length field and
corrupt the sidecar. A real manifest is a few KB, so this never trips.

* fix(ec_bitrot): cap -ec.bitrotBlockSizeMB at 64 MiB

The block size becomes the per-shard scratch buffer the scrub/backfill path
allocates, so an over-large value (e.g. 1 GiB) is a memory hazard per concurrent
scrub worker. Lower the upper bound from 1024 to 64 MiB.

* fix(ec_bitrot): add -ecUnsafeIgnoreSidecar to weed tool fix -ecx

The -ecx recovery path reconstructs missing shards via RebuildEcFilesWithContext,
which fails closed on a malformed/stale .ecsum. Without an override flag an
operator could not complete the rebuild without manually deleting the sidecar.
Expose -ecUnsafeIgnoreSidecar (default false) and thread it through.

* fix(ec_bitrot): bound sidecar payload with a direct int constant; drop readFull

Guard len(payload) against a plain int constant (1 GiB) before the allocation
instead of a uint64 MaxUint32 compare, so the allocation-size value is provably
bounded (clears the CodeQL overflow alert) and the math import is no longer
needed. Inline os.File.ReadAt with io.EOF handling in verifyShardFileBlocks and
remove the now-redundant readFull helper (os.File.ReadAt fills the slice or
errors).

* test(ec_bitrot): use slices.Contains instead of a hand-rolled containsU32

* refactor(ec): fold the EcFiles WithContext variants into the base functions

RebuildEcFiles now takes the *ECContext directly (nil => derive from .vif as
before) and WriteEcFiles takes it too (nil => default), removing the parallel
RebuildEcFilesWithContext / WriteEcFilesWithContext names. Callers that had an
explicit context drop the WithContext suffix; the default-context callers pass
nil. No behavior change.

* refactor(ec): pass BackgroundECContext instead of nil to Write/RebuildEcFiles

Add a non-nil BackgroundECContext placeholder (analogous to context.Background())
and have callers with no specific layout pass it instead of a nil *ECContext.
WriteEcFiles resolves a zero/background context to the default ratio and
RebuildEcFiles resolves it from the .vif, so behavior is unchanged.

* fix(ec_bitrot): make BackgroundECContext a func; RebuildEcFiles fails closed on bad .vif

- BackgroundECContext is now a function returning a fresh *ECContext, so callers
  cannot mutate a shared singleton or race on it (and it mirrors context.Background,
  which is also a function).
- RebuildEcFiles now propagates the MaybeLoadVolumeInfo error: a present-but-
  unreadable .vif fails closed instead of silently rebuilding with the default
  ratio (which would corrupt a custom-ratio volume). Pass an explicit ctx to override.
2026-05-31 18:52:44 -07:00

538 lines
21 KiB
Go

package command
import (
"context"
"crypto/tls"
"net/http"
httppprof "net/http/pprof"
"os"
"runtime/pprof"
"strconv"
"strings"
"time"
"github.com/spf13/viper"
"google.golang.org/grpc"
"google.golang.org/grpc/reflection"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb"
"github.com/seaweedfs/seaweedfs/weed/pb/volume_server_pb"
"github.com/seaweedfs/seaweedfs/weed/security"
weed_server "github.com/seaweedfs/seaweedfs/weed/server"
"github.com/seaweedfs/seaweedfs/weed/server/constants"
stats_collect "github.com/seaweedfs/seaweedfs/weed/stats"
"github.com/seaweedfs/seaweedfs/weed/storage"
"github.com/seaweedfs/seaweedfs/weed/storage/erasure_coding"
"github.com/seaweedfs/seaweedfs/weed/storage/types"
"github.com/seaweedfs/seaweedfs/weed/util"
"github.com/seaweedfs/seaweedfs/weed/util/grace"
"github.com/seaweedfs/seaweedfs/weed/util/httpdown"
"github.com/seaweedfs/seaweedfs/weed/util/version"
)
var (
v VolumeServerOptions
)
type VolumeServerOptions struct {
port *int
portGrpc *int
publicPort *int
folders []string
folderMaxLimits []int32
idxFolder *string
ip *string
id *string
publicUrl *string
bindIp *string
mastersString *string
mserverString *string // deprecated, for backward compatibility
masters []pb.ServerAddress
idleConnectionTimeout *int
dataCenter *string
rack *string
whiteList []string
indexType *string
diskType *string
tags *string
fixJpgOrientation *bool
readMode *string
cpuProfile *string
memProfile *string
compactionMBPerSecond *int
maintenanceMBPerSecond *int
fileSizeLimitMB *int
concurrentUploadLimitMB *int
concurrentDownloadLimitMB *int
pprof *bool
preStopSeconds *int
metricsHttpPort *int
metricsHttpIp *string
// pulseSeconds *int
inflightUploadDataTimeout *time.Duration
inflightDownloadDataTimeout *time.Duration
hasSlowRead *bool
readBufferSizeMB *int
ldbTimeout *int64
allowUntrustedRemoteEndpoints *bool
debug *bool
debugPort *int
// shutdownCtx, when non-nil, tells startVolumeServer to shut down once the
// ctx is cancelled. Used by integration tests and by weed mini; nil for
// standalone weed volume.
shutdownCtx context.Context
}
func init() {
cmdVolume.Run = runVolume // break init cycle
v.port = cmdVolume.Flag.Int("port", 8080, "http listen port")
v.portGrpc = cmdVolume.Flag.Int("port.grpc", 0, "grpc listen port")
v.publicPort = cmdVolume.Flag.Int("port.public", 0, "port opened to public")
v.ip = cmdVolume.Flag.String("ip", util.DetectedHostAddress(), "ip or server name, also used as identifier")
v.id = cmdVolume.Flag.String("id", "", "volume server id. If empty, default to ip:port")
v.publicUrl = cmdVolume.Flag.String("publicUrl", "", "Publicly accessible address")
v.bindIp = cmdVolume.Flag.String("ip.bind", "", "ip address to bind to. If empty, default to same as -ip option.")
v.mastersString = cmdVolume.Flag.String("master", "localhost:9333", "comma-separated master servers")
v.mserverString = cmdVolume.Flag.String("mserver", "", "comma-separated master servers (deprecated, use -master instead)")
v.preStopSeconds = cmdVolume.Flag.Int("preStopSeconds", 10, "number of seconds between stop send heartbeats and stop volume server")
// v.pulseSeconds = cmdVolume.Flag.Int("pulseSeconds", 5, "number of seconds between heartbeats, must be smaller than or equal to the master's setting")
v.idleConnectionTimeout = cmdVolume.Flag.Int("idleTimeout", 30, "connection idle seconds")
v.dataCenter = cmdVolume.Flag.String("dataCenter", "", "current volume server's data center name")
v.rack = cmdVolume.Flag.String("rack", "", "current volume server's rack name")
v.indexType = cmdVolume.Flag.String("index", "memory", "Choose [memory|leveldb|leveldbMedium|leveldbLarge] mode for memory~performance balance.")
v.diskType = cmdVolume.Flag.String("disk", "", "[hdd|ssd|<tag>] hard drive or solid state drive or any tag")
v.tags = cmdVolume.Flag.String("tags", "", "comma-separated tag groups per data dir; each group uses ':' (e.g. fast:ssd,archive)")
v.fixJpgOrientation = cmdVolume.Flag.Bool("images.fix.orientation", false, "Adjust jpg orientation when uploading.")
v.readMode = cmdVolume.Flag.String("readMode", "proxy", "[local|proxy|redirect] how to deal with non-local volume: 'not found|proxy to remote node|redirect volume location'.")
v.cpuProfile = cmdVolume.Flag.String("cpuprofile", "", "cpu profile output file")
v.memProfile = cmdVolume.Flag.String("memprofile", "", "memory profile output file")
v.compactionMBPerSecond = cmdVolume.Flag.Int("compactionMBps", 0, "limit background compaction or copying speed in mega bytes per second")
v.maintenanceMBPerSecond = cmdVolume.Flag.Int("maintenanceMBps", 0, "limit maintenance (replication / balance) IO rate in MB/s. Unset is 0, no limitation.")
v.fileSizeLimitMB = cmdVolume.Flag.Int("fileSizeLimitMB", 256, "limit file size to avoid out of memory")
v.ldbTimeout = cmdVolume.Flag.Int64("index.leveldbTimeout", 0, "alive time for leveldb (default to 0). If leveldb of volume is not accessed in ldbTimeout hours, it will be off loaded to reduce opened files and memory consumption.")
v.concurrentUploadLimitMB = cmdVolume.Flag.Int("concurrentUploadLimitMB", 0, "limit total concurrent upload size, 0 means unlimited")
v.concurrentDownloadLimitMB = cmdVolume.Flag.Int("concurrentDownloadLimitMB", 0, "limit total concurrent download size, 0 means unlimited")
v.pprof = cmdVolume.Flag.Bool("pprof", false, "enable pprof http handlers. precludes -memprofile and -cpuprofile")
v.metricsHttpPort = cmdVolume.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
v.metricsHttpIp = cmdVolume.Flag.String("metricsIp", "", "metrics listen ip. If empty, default to same as -ip.bind option.")
v.idxFolder = cmdVolume.Flag.String("dir.idx", "", "directory to store .idx files")
v.inflightUploadDataTimeout = cmdVolume.Flag.Duration("inflightUploadDataTimeout", 60*time.Second, "inflight upload data wait timeout of volume servers")
v.inflightDownloadDataTimeout = cmdVolume.Flag.Duration("inflightDownloadDataTimeout", 60*time.Second, "inflight download data wait timeout of volume servers")
v.hasSlowRead = cmdVolume.Flag.Bool("hasSlowRead", true, "<experimental> if true, this prevents slow reads from blocking other requests, but large file read P99 latency will increase.")
v.readBufferSizeMB = cmdVolume.Flag.Int("readBufferSizeMB", 4, "<experimental> larger values can optimize query performance but will increase some memory usage,Use with hasSlowRead normally.")
v.allowUntrustedRemoteEndpoints = cmdVolume.Flag.Bool("volume.allowUntrustedRemoteEndpoints", false, "if true, FetchAndWriteNeedle accepts arbitrary remote S3 endpoints including loopback / link-local hosts. Default rejects internal / metadata endpoints.")
v.debug = cmdVolume.Flag.Bool("debug", false, "serves runtime profiling data via pprof on the port specified by -debug.port")
v.debugPort = cmdVolume.Flag.Int("debug.port", 6060, "http port for debugging")
}
var cmdVolume = &Command{
UsageLine: "volume -port=8080 -dir=/tmp -max=5 -ip=server_name -master=localhost:9333",
Short: "start a volume server",
Long: `start a volume server to provide storage spaces
`,
}
var (
volumeFolders = cmdVolume.Flag.String("dir", os.TempDir(), "directories to store data files. dir[,dir]...")
maxVolumeCounts = cmdVolume.Flag.String("max", "8", "maximum numbers of volumes, count[,count]... If set to zero, the limit will be auto configured as free disk space divided by volume size.")
volumeWhiteListOption = cmdVolume.Flag.String("whiteList", "", "comma separated Ip addresses having write permission. No limit if empty.")
minFreeSpacePercent = cmdVolume.Flag.String("minFreeSpacePercent", "1", "minimum free disk space (default to 1%). Low disk space will mark all volumes as ReadOnly (deprecated, use minFreeSpace instead).")
minFreeSpace = cmdVolume.Flag.String("minFreeSpace", "", "min free disk space (value<=100 as percentage like 1, other as human readable bytes, like 10GiB). Low disk space will mark all volumes as ReadOnly.")
ecBitrotChecksum = cmdVolume.Flag.Bool("ec.bitrotChecksum", true, "write a bitrot checksum sidecar (.ecsum) when generating EC shards, enabling silent-corruption detection on scrub/rebuild")
ecBitrotBlockSizeMB = cmdVolume.Flag.Int("ec.bitrotBlockSizeMB", 16, "EC bitrot checksum block granularity in MiB; must be a power of two and at least 1")
)
func runVolume(cmd *Command, args []string) bool {
if *v.debug {
grace.StartDebugServer(*v.debugPort)
}
util.LoadSecurityConfiguration()
// If --pprof is set we assume the caller wants to be able to collect
// cpu and memory profiles via go tool pprof
if !*v.pprof {
*v.cpuProfile = util.ResolvePath(*v.cpuProfile)
*v.memProfile = util.ResolvePath(*v.memProfile)
grace.SetupProfiling(*v.cpuProfile, *v.memProfile)
}
switch {
case *v.metricsHttpIp != "":
// noting to do, use v.metricsHttpIp
case *v.bindIp != "":
*v.metricsHttpIp = *v.bindIp
case *v.ip != "":
*v.metricsHttpIp = *v.ip
}
go stats_collect.StartMetricsServer(*v.metricsHttpIp, *v.metricsHttpPort)
// Backward compatibility: if -mserver is provided, use it
if *v.mserverString != "" {
*v.mastersString = *v.mserverString
}
// Apply EC bitrot checksum settings.
erasure_coding.BitrotProtectionEnabled = *ecBitrotChecksum
// Validate the block size before multiplying so an absurd MiB value cannot
// overflow int64 and slip a bogus size past the power-of-two check. The
// block size also becomes the per-shard scratch buffer the scrub/backfill
// path allocates, so the upper bound caps that allocation (64 MiB per
// concurrent scrub worker) and keeps a typo from taking the server down.
const maxBitrotBlockSizeMB = 64
if mb := *ecBitrotBlockSizeMB; mb >= 1 && mb <= maxBitrotBlockSizeMB {
if blockSize := int64(mb) * 1024 * 1024; blockSize&(blockSize-1) == 0 {
erasure_coding.BitrotBlockSize = blockSize
} else if mb != 16 {
glog.Warningf("ignoring invalid -ec.bitrotBlockSizeMB=%d (must be a power of two); using %d MiB",
mb, erasure_coding.BitrotBlockSize/(1024*1024))
}
} else if *ecBitrotBlockSizeMB != 16 {
glog.Warningf("ignoring out-of-range -ec.bitrotBlockSizeMB=%d (must be a power of two in [1, %d] MiB); using %d MiB",
*ecBitrotBlockSizeMB, maxBitrotBlockSizeMB, erasure_coding.BitrotBlockSize/(1024*1024))
}
minFreeSpaces := util.MustParseMinFreeSpace(*minFreeSpace, *minFreeSpacePercent)
v.masters = pb.ServerAddresses(*v.mastersString).ToAddresses()
v.startVolumeServer(*volumeFolders, *maxVolumeCounts, *volumeWhiteListOption, minFreeSpaces)
return true
}
func (v VolumeServerOptions) startVolumeServer(volumeFolders, maxVolumeCounts, volumeWhiteListOption string, minFreeSpaces []util.MinFreeSpace) {
// Set multiple folders and each folder's max volume count limit'
v.folders = strings.Split(volumeFolders, ",")
for i, folder := range v.folders {
v.folders[i] = util.ResolvePath(folder)
if err := util.TestFolderWritable(v.folders[i]); err != nil {
glog.Fatalf("Check Data Folder(-dir) Writable %s : %s", v.folders[i], err)
}
}
// set max
maxCountStrings := strings.Split(maxVolumeCounts, ",")
for _, maxString := range maxCountStrings {
if max, e := strconv.ParseInt(maxString, 10, 64); e == nil {
v.folderMaxLimits = append(v.folderMaxLimits, int32(max))
} else {
glog.Fatalf("The max specified in -max not a valid number %s", maxString)
}
}
if len(v.folderMaxLimits) == 1 && len(v.folders) > 1 {
for i := 0; i < len(v.folders)-1; i++ {
v.folderMaxLimits = append(v.folderMaxLimits, v.folderMaxLimits[0])
}
}
if len(v.folders) != len(v.folderMaxLimits) {
glog.Fatalf("%d directories by -dir, but only %d max is set by -max", len(v.folders), len(v.folderMaxLimits))
}
if len(minFreeSpaces) == 1 && len(v.folders) > 1 {
for i := 0; i < len(v.folders)-1; i++ {
minFreeSpaces = append(minFreeSpaces, minFreeSpaces[0])
}
}
if len(v.folders) != len(minFreeSpaces) {
glog.Fatalf("%d directories by -dir, but only %d minFreeSpacePercent is set by -minFreeSpacePercent", len(v.folders), len(minFreeSpaces))
}
// set disk types
var diskTypes []types.DiskType
diskTypeStrings := strings.Split(*v.diskType, ",")
for _, diskTypeString := range diskTypeStrings {
diskTypes = append(diskTypes, types.ToDiskType(diskTypeString))
}
if len(diskTypes) == 1 && len(v.folders) > 1 {
for i := 0; i < len(v.folders)-1; i++ {
diskTypes = append(diskTypes, diskTypes[0])
}
}
if len(v.folders) != len(diskTypes) {
glog.Fatalf("%d directories by -dir, but only %d disk types is set by -disk", len(v.folders), len(diskTypes))
}
var tagsArg string
if v.tags != nil {
tagsArg = *v.tags
}
folderTags := parseVolumeTags(tagsArg, len(v.folders))
// security related white list configuration
v.whiteList = util.StringSplit(volumeWhiteListOption, ",")
if *v.ip == "" {
*v.ip = util.DetectedHostAddress()
glog.V(0).Infof("detected volume server ip address: %v", *v.ip)
}
if *v.bindIp == "" {
*v.bindIp = *v.ip
}
if *v.publicPort == 0 {
*v.publicPort = *v.port
}
if *v.portGrpc == 0 {
*v.portGrpc = 10000 + *v.port
}
if *v.publicUrl == "" {
*v.publicUrl = util.JoinHostPort(*v.ip, *v.publicPort)
}
volumeMux := http.NewServeMux()
publicVolumeMux := volumeMux
if v.isSeparatedPublicPort() {
publicVolumeMux = http.NewServeMux()
}
if *v.pprof {
volumeMux.HandleFunc("/debug/pprof/", httppprof.Index)
volumeMux.HandleFunc("/debug/pprof/cmdline", httppprof.Cmdline)
volumeMux.HandleFunc("/debug/pprof/profile", httppprof.Profile)
volumeMux.HandleFunc("/debug/pprof/symbol", httppprof.Symbol)
volumeMux.HandleFunc("/debug/pprof/trace", httppprof.Trace)
}
volumeNeedleMapKind := storage.NeedleMapInMemory
switch *v.indexType {
case "leveldb":
volumeNeedleMapKind = storage.NeedleMapLevelDb
case "leveldbMedium":
volumeNeedleMapKind = storage.NeedleMapLevelDbMedium
case "leveldbLarge":
volumeNeedleMapKind = storage.NeedleMapLevelDbLarge
}
// Determine volume server ID: if not specified, use ip:port
volumeServerId := util.GetVolumeServerId(*v.id, *v.ip, *v.port)
volumeServer := weed_server.NewVolumeServer(volumeMux, publicVolumeMux,
*v.ip, *v.port, *v.portGrpc, *v.publicUrl, volumeServerId,
v.folders, v.folderMaxLimits, minFreeSpaces, diskTypes, folderTags,
util.ResolvePath(*v.idxFolder),
volumeNeedleMapKind,
v.masters, constants.VolumePulsePeriod, *v.dataCenter, *v.rack,
v.whiteList,
*v.fixJpgOrientation, *v.readMode,
*v.compactionMBPerSecond,
*v.maintenanceMBPerSecond,
*v.fileSizeLimitMB,
int64(*v.concurrentUploadLimitMB)*1024*1024,
int64(*v.concurrentDownloadLimitMB)*1024*1024,
*v.inflightUploadDataTimeout,
*v.inflightDownloadDataTimeout,
*v.hasSlowRead,
*v.readBufferSizeMB,
*v.ldbTimeout,
*v.allowUntrustedRemoteEndpoints,
)
// starting grpc server
grpcS := v.startGrpcService(volumeServer)
// starting public http server
var publicHttpDown httpdown.Server
if v.isSeparatedPublicPort() {
publicHttpDown = v.startPublicHttpService(publicVolumeMux)
if nil == publicHttpDown {
glog.Fatalf("start public http service failed")
}
}
// starting the cluster http server
clusterHttpServer, closeCert := v.startClusterHttpService(volumeMux)
grace.OnReload(volumeServer.LoadNewVolumes)
grace.OnReload(volumeServer.Reload)
stopChan := make(chan bool)
grace.OnInterrupt(func() {
glog.Infof("volume server has been killed")
// Stop heartbeats
if !volumeServer.StopHeartbeat() {
volumeServer.SetStopping()
glog.V(0).Infof("stop send heartbeat and wait %d seconds until shutdown ...", *v.preStopSeconds)
time.Sleep(time.Duration(*v.preStopSeconds) * time.Second)
}
shutdown(publicHttpDown, clusterHttpServer, grpcS, volumeServer)
if closeCert != nil {
closeCert()
}
stopChan <- true
})
if v.shutdownCtx != nil {
select {
case <-stopChan:
case <-v.shutdownCtx.Done():
shutdown(publicHttpDown, clusterHttpServer, grpcS, volumeServer)
if closeCert != nil {
closeCert()
}
}
} else {
select {
case <-stopChan:
}
}
}
func parseVolumeTags(tagsArg string, folderCount int) [][]string {
if folderCount <= 0 {
return nil
}
tagEntries := []string{}
if strings.TrimSpace(tagsArg) != "" {
tagEntries = strings.Split(tagsArg, ",")
}
folderTags := make([][]string, folderCount)
// If exactly one tag entry provided, replicate it to all folders
if len(tagEntries) == 1 {
normalized := util.NormalizeTagList(strings.Split(tagEntries[0], ":"))
for i := 0; i < folderCount; i++ {
folderTags[i] = append([]string(nil), normalized...)
}
} else {
// Otherwise, assign tags to folders that have explicit entries
for i := 0; i < folderCount; i++ {
if i < len(tagEntries) {
folderTags[i] = util.NormalizeTagList(strings.Split(tagEntries[i], ":"))
} else {
// Initialize remaining folders with empty tag slice
folderTags[i] = []string{}
}
}
}
return folderTags
}
func shutdown(publicHttpDown httpdown.Server, clusterHttpServer httpdown.Server, grpcS *grpc.Server, volumeServer *weed_server.VolumeServer) {
// firstly, stop the public http service to prevent from receiving new user request
if nil != publicHttpDown {
glog.V(0).Infof("stop public http server ... ")
if err := publicHttpDown.Stop(); err != nil {
glog.Warningf("stop the public http server failed, %v", err)
}
}
glog.V(0).Infof("graceful stop cluster http server ... ")
if err := clusterHttpServer.Stop(); err != nil {
glog.Warningf("stop the cluster http server failed, %v", err)
}
glog.V(0).Infof("graceful stop gRPC ...")
grpcS.GracefulStop()
volumeServer.Shutdown()
pprof.StopCPUProfile()
}
// check whether configure the public port
func (v VolumeServerOptions) isSeparatedPublicPort() bool {
return *v.publicPort != *v.port
}
func (v VolumeServerOptions) startGrpcService(vs volume_server_pb.VolumeServerServer) *grpc.Server {
grpcPort := *v.portGrpc
grpcL, err := util.NewListener(util.JoinHostPort(*v.bindIp, grpcPort), 0)
if err != nil {
glog.Fatalf("failed to listen on grpc port %d: %v", grpcPort, err)
}
grpcS := pb.NewGrpcServer(security.LoadServerTLS(util.GetViper(), "grpc.volume"))
volume_server_pb.RegisterVolumeServerServer(grpcS, vs)
reflection.Register(grpcS)
go func() {
if err := grpcS.Serve(grpcL); err != nil {
glog.Fatalf("start gRPC service failed, %s", err)
}
}()
pb.ServeGrpcOnLocalSocket(grpcS, grpcPort)
return grpcS
}
func (v VolumeServerOptions) startPublicHttpService(handler http.Handler) httpdown.Server {
publicListeningAddress := util.JoinHostPort(*v.bindIp, *v.publicPort)
glog.V(0).Infoln("Start Seaweed volume server", version.Version(), "public at", publicListeningAddress)
publicListener, e := util.NewListener(publicListeningAddress, time.Duration(*v.idleConnectionTimeout)*time.Second)
if e != nil {
glog.Fatalf("Volume server listener error:%v", e)
}
pubHttp := httpdown.HTTP{StopTimeout: 5 * time.Minute, KillTimeout: 5 * time.Minute}
publicHttpDown := pubHttp.Serve(&http.Server{Handler: handler}, publicListener)
go func() {
if err := publicHttpDown.Wait(); err != nil {
glog.Errorf("public http down wait failed, %v", err)
}
}()
return publicHttpDown
}
// startClusterHttpService starts the volume cluster HTTP server and
// returns it along with a close func for the cert reloader's refresh
// goroutine (nil when HTTPS is disabled). The caller is responsible
// for invoking the close func on every shutdown path — both the
// SIGTERM/grace.OnInterrupt path and the shutdownCtx path used by
// mini/integration tests.
func (v VolumeServerOptions) startClusterHttpService(handler http.Handler) (httpdown.Server, func()) {
var (
certFile, keyFile string
)
if viper.GetString("https.volume.key") != "" {
certFile = viper.GetString("https.volume.cert")
keyFile = viper.GetString("https.volume.key")
}
listeningAddress := util.JoinHostPort(*v.bindIp, *v.port)
glog.V(0).Infof("Start Seaweed volume server %s at %s", version.Version(), listeningAddress)
listener, e := util.NewListener(listeningAddress, time.Duration(*v.idleConnectionTimeout)*time.Second)
if e != nil {
glog.Fatalf("Volume server listener error:%v", e)
}
httpDown := httpdown.HTTP{
KillTimeout: time.Minute,
StopTimeout: 30 * time.Second,
}
httpS := &http.Server{Handler: handler}
if viper.GetString("https.volume.ca") != "" {
clientCertFile := viper.GetString("https.volume.ca")
httpS.TLSConfig = security.LoadClientTLSHTTP(clientCertFile)
if err := security.FixTlsConfig(util.GetViper(), httpS.TLSConfig); err != nil {
glog.Fatalf("Could not fix TLS config: %v", err)
}
}
var closeCert func()
if certFile != "" && keyFile != "" {
getCert, certProvider, err := security.NewReloadingServerCertificate(certFile, keyFile)
if err != nil {
glog.Fatalf("Volume server failed to load TLS certificate: %v", err)
}
closeCert = certProvider.Close
if httpS.TLSConfig == nil {
httpS.TLSConfig = &tls.Config{}
}
httpS.TLSConfig.GetCertificate = getCert
}
clusterHttpServer := httpDown.Serve(httpS, listener)
go func() {
if e := clusterHttpServer.Wait(); e != nil {
glog.Fatalf("Volume server fail to serve: %v", e)
}
}()
return clusterHttpServer, closeCert
}