Github-Mirror/seaweedfs
1
0 Fork 0
mirror of https://github.com/seaweedfs/seaweedfs.git synced 2026-10-30 00:40:44 +01:00
Code Issues Packages Projects Releases Wiki Activity
Files
fffdb7cc754d34e95583657f0c1342798b127a69
seaweedfs/weed/command/server.go
T
Chris LuandDevin 18691e7b5b server: keep the first write on a fresh weed server from stalling (#12149) 
* master: end the post-mint allocation hold on the embedded reporter's clean inventory

DeferAllocation armed a warmup-length hold whenever the leader minted a
fresh topology id, and on weed server the hold ran the whole window: the
embedded volume server's join kept refreshing the quiet evidence clock, so
the two-pulse early release still cost the first write 10-15s on every
fresh start. weed mini already releases on its single pinned reporter's
first inventory, but pinning refuses outside volume servers, which weed
server must keep admitting.

Add a non-exclusive variant: the topology names one expected reporter
whose first clean full inventory lifts the hold, while other reporters
stay admitted and still feed prior-state evidence and the topology-id
freeze. A single-master weed server points it at its embedded volume
server — the one reporter guaranteed to arrive quickly; a veteran's
heartbeat carries prior state and keeps the window through the regular
checks, so a wiped -dir with surviving external volume servers still
trips the freeze on first report. Restarted clusters are untouched: a
recovered topology id means no mint and no hold.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* filer: announce on the startup peer scan for embedded single-master setups

A fresh filer whose startup peer scan came back empty re-polled for the
full confirm window before registering on the master, so the lock ring
had no member and routed traffic such as S3 PUTs spun on "no lock server
found" for ~10s on every fresh weed server / weed mini start.

For setups that start with exactly one filer the scan is already final:
a single-master weed server and weed mini announce as soon as it
completes. A late external filer still joins and replays through the
normal path — the confirm window only ever caught peers arriving inside
it, and the residual risk mirrors the accepted volume-side trade-off: an
embedded filer announcing a few seconds earlier can own the lock ring
while a surviving external filer's backlog is still on its way.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* volume: retry a failed heartbeat promptly until first registration

A heartbeat stream refused before the leader is known — the common race
when a volume server starts while an election is still running — always
fell back to sleeping out the full pulse, so a fresh volume's first
registration could wait ~5s behind a leader elected a second later.

The peer refresh already identifies the leader during the wait, so cap
the retry pause at a second once a leader is named; while no leader is
known the same short retry applies until the first stream is confirmed,
since an unregistered server has nothing else to do. Once the server was
registered the pulse backoff is unchanged, so a lost leader still paces
reconnects the way it did before.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* volume: retry a failed heartbeat promptly until first registration (rust)

Mirror the Go heartbeat retry: cap the post-refresh wait at a second when
a leader is named, and keep the short retry until the first stream is
confirmed so a fresh volume registers promptly when it races an election.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* master: leave admitted reporters a confirm pulse when the embedded report ends the hold

The single-master weed server releases its post-mint allocation hold on
the embedded volume server's clean report, but outside volume servers
stay admitted and one may still carry prior cluster state — an inventory
from the embedded server alone cannot prove none is due. Clearing the
hold outright let a rebuilt master re-issue a volume id a reconnecting
veteran still owned.

Collapse the released hold to one heartbeat interval instead: late
reporters land their evidence or a topology-id freeze inside the pulse,
and a re-mint re-arms that pulse rather than the full warmup. The pinned
single-reporter release (weed mini) still drops the hold outright, since
no other reporter can be admitted.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* filer: keep the empty-membership confirm on embedded single-master weed server

A single-master weed server still admits external filers, so an empty
startup peer scan is not final: after a master restart the existing
filers simply have not reconnected yet. Announcing on the scan let the
embedded filer take lock-ring ownership and serve writes that a peer's
replayed backlog could then overwrite.

Only weed mini keeps the skip, where the embedded filer is the only
member its dev cluster can have. The confirm window only delays the
master listing for a fresh store; the filer serves while it runs.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* volume: tolerate an expired heartbeat retry deadline (rust)

refresh_leader_from_peers can consume the whole pulse budget, leaving
retry_at in the past; subtracting a later Instant panics and kills the
heartbeat task instead of retrying. Match the Go side, where
time.Until of a past deadline just sleeps zero.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* volume: retry the seed list without a second pulse sleep (rust)

The error path already slept its retry delay; an extra full pulse after
losing the round delayed the next attempt past what Go's heartbeat()
does, which paces retries only inside the per-master attempt.

* master: keep a configured allocation hold shorter than the confirm pulse

Once the embedded report lands, the collapsed hold replaced the
configured deferral with a full pulse even when the operator set a
shorter one, including zero (disabled). Cap at the pulse instead.

* filer: end the empty-membership confirm early on a fresh-minted embedded master

On a fresh single-master weed server the announce gate still held the
filer for the full confirm window, so the master's lock ring stayed
empty and the first write retried 'no lock server found' for ~7s on
top of the collapsed allocation pulse. A fresh TopologyId mint proves
the master's persisted state was empty, making the empty peer scan
final; skip the rest of the confirm only then, and keep it whenever
the master recovers a prior id.

* filer: keep a short reconnect window on a fresh-minted embedded master

A fresh TopologyId mint proves the master's persisted state was empty,
not that no external filer survived its loss — one can still be
mid-reconnect on its keepalive cycle when the confirm starts. Ending
the empty-scan confirm on the first poll let the fresh filer register
and serve routed reads before that peer was found and backfilled, so
existing files could read as missing. The fresh-mint path now shortens
the confirm to a few keepalive cycles instead of skipping it: a
reconnecting filer still lands in time, while a genuinely fresh cluster
announces after ~3s rather than ~10s.

* volume: skip the heartbeat loop when no seed masters are configured

An empty seed list made the outer loop spin with no await between
passes and no path to observe the stop flag. A masterless volume server
has nothing to register with, so the heartbeat exits after a warning.

* volume: skip the heartbeat loop when no master addresses are given (rust)

Mirrors the Go heartbeat: an empty seed list spun the outer loop with
no await and no shutdown checks. Exit after a warning so library
callers get the same behavior as the binary path that validates the
list up front.

* filer: re-check the fresh-mint hint on each confirm pass

The topology id is minted after raft election, which can land after
the empty-scan confirm has already started — a one-time check at entry
missed the mint and held the full window. Evaluate the hint on every
poll and shorten the confirm from the first observed mint, still
capped by the normal confirm deadline.

* master: mark a fresh topology mint only when the volume bound is zero

Recovery restores the max volume id unconditionally but the topology id
only when present, so an upgraded or partially rebuilt master can mint a
new id while still holding prior volume state. Marking fresh there would
let the embedded filer shortcut its peer confirm on a cluster that is
not empty; only mark the mint fresh when no bound survived.

* filer: drop the fresh-mint shortcut once a peer has been observed

---------

Co-authored-by: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-10-28 08:08:26 -07:00

564 lines
35 KiB
Go
Raw Blame History

package command
import (
"crypto/tls"
"fmt"
"math"
"net/http"
"os"
"strings"
"sync/atomic"
"time"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb"
weed_server "github.com/seaweedfs/seaweedfs/weed/server"
stats_collect "github.com/seaweedfs/seaweedfs/weed/stats"
"github.com/seaweedfs/seaweedfs/weed/util"
flag "github.com/seaweedfs/seaweedfs/weed/util/fla9"
"github.com/seaweedfs/seaweedfs/weed/util/grace"
)
type ServerOptions struct {
cpuprofile *string
memprofile *string
debug *bool
debugPort *int
v VolumeServerOptions
}
var (
serverOptions ServerOptions
masterOptions MasterOptions
filerOptions FilerOptions
s3Options S3Options
sftpOptions SftpOptions
iamOptions IamOptions
webdavOptions WebDavOption
mqBrokerOptions MessageQueueBrokerOptions
mqAgentServerOptions MessageQueueAgentOptions
)
func init() {
cmdServer.Run = runServer // break init cycle
}
var cmdServer = &Command{
UsageLine: "server -dir=/tmp -volume.max=5 -ip=server_name",
Short: "start a master server, a volume server, and optionally a filer and a S3 gateway",
Long: `start both a volume server to provide storage spaces
and a master server to provide volume=>location mapping service and sequence number of file ids
This is provided as a convenient way to start both volume server and master server.
The servers acts exactly the same as starting them separately.
So other volume servers can connect to this master server also.
Optionally, a filer server can be started.
Also optionally, a S3 gateway can be started.
`,
}
var (
serverIp = cmdServer.Flag.String("ip", util.DetectedHostAddress(), "ip or server name, also used as identifier")
serverBindIp = cmdServer.Flag.String("ip.bind", "", "ip address to bind to. If empty, default to same as -ip option.")
serverTimeout = cmdServer.Flag.Int("idleTimeout", 30, "connection idle seconds")
serverDataCenter = cmdServer.Flag.String("dataCenter", "", "current volume server's data center name")
serverRack = cmdServer.Flag.String("rack", "", "current volume server's rack name")
serverWhiteListOption = cmdServer.Flag.String("whiteList", "", "comma separated Ip addresses having write permission. No limit if empty.")
serverDisableHttp = cmdServer.Flag.Bool("disableHttp", false, "disable http requests, only gRPC operations are allowed.")
serverIamConfig = cmdServer.Flag.String("iam.config", "", "path to the advanced IAM config file for S3. An alias for -s3.iam.config, but with lower priority.")
volumeDataFolders = cmdServer.Flag.String("dir", os.TempDir(), "directories to store data files. dir[,dir]...")
volumeMaxDataVolumeCounts = cmdServer.Flag.String("volume.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.")
volumeMinFreeSpacePercent = cmdServer.Flag.String("volume.minFreeSpacePercent", "1", "minimum free disk space (default to 1%). Low disk space will mark all volumes as ReadOnly (deprecated, use minFreeSpace instead).")
volumeMinFreeSpace = cmdServer.Flag.String("volume.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.")
serverMetricsHttpPort = cmdServer.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
serverMetricsHttpIp = cmdServer.Flag.String("metricsIp", "", "metrics listen ip. If empty, default to same as -ip.bind option.")
// pulseSeconds = cmdServer.Flag.Int("pulseSeconds", 5, "number of seconds between heartbeats")
isStartingMasterServer = cmdServer.Flag.Bool("master", true, "whether to start master server")
isStartingVolumeServer = cmdServer.Flag.Bool("volume", true, "whether to start volume server")
isStartingFiler = cmdServer.Flag.Bool("filer", false, "whether to start filer")
isStartingS3 = cmdServer.Flag.Bool("s3", false, "whether to start S3 gateway")
isStartingSftp = cmdServer.Flag.Bool("sftp", false, "whether to start Sftp server")
isStartingIam = cmdServer.Flag.Bool("iam", false, "whether to start IAM service")
isStartingWebDav = cmdServer.Flag.Bool("webdav", false, "whether to start WebDAV gateway")
isStartingMqBroker = cmdServer.Flag.Bool("mq.broker", false, "whether to start message queue broker")
isStartingMqAgent = cmdServer.Flag.Bool("mq.agent", false, "whether to start message queue agent")
False = false
)
func init() {
serverOptions.cpuprofile = cmdServer.Flag.String("cpuprofile", "", "cpu profile output file")
serverOptions.memprofile = cmdServer.Flag.String("memprofile", "", "memory profile output file")
serverOptions.debug = cmdServer.Flag.Bool("debug", false, "serves runtime profiling data, e.g., http://localhost:6060/debug/pprof/goroutine?debug=2")
serverOptions.debugPort = cmdServer.Flag.Int("debug.port", 6060, "http port for debugging")
masterOptions.port = cmdServer.Flag.Int("master.port", 9333, "master server http listen port")
masterOptions.portGrpc = cmdServer.Flag.Int("master.port.grpc", 0, "master server grpc listen port")
masterOptions.metaFolder = cmdServer.Flag.String("master.dir", "", "data directory to store meta data, default to same as -dir specified")
masterOptions.peers = cmdServer.Flag.String("master.peers", "", "all master nodes in comma separated ip:masterPort list")
masterOptions.volumeSizeLimitMB = cmdServer.Flag.Uint("master.volumeSizeLimitMB", util.DefaultVolumeSizeLimitMB, "Master stops directing writes to oversized volumes.")
masterOptions.volumePreallocate = cmdServer.Flag.Bool("master.volumePreallocate", false, "Preallocate disk space for volumes.")
masterOptions.maxParallelVacuumPerServer = cmdServer.Flag.Int("master.maxParallelVacuumPerServer", 1, "maximum number of volumes to vacuum in parallel on one volume server; 0 for no limit (deprecated, use master.vacuum.max_parallel_per_server in master.toml)")
masterOptions.vacuumIntervalSeconds = cmdServer.Flag.Int("master.vacuumIntervalSeconds", 840, "seconds between automatic vacuum sweeps")
masterOptions.vacuumDeleteEmptyAfterSeconds = cmdServer.Flag.Int("master.vacuumDeleteEmptyAfterSeconds", 0, "automatic sweep deletes volume copies that stay empty this many seconds; 0 disables")
masterOptions.defaultReplication = cmdServer.Flag.String("master.defaultReplication", "", "Default replication type if not specified.")
masterOptions.garbageThreshold = cmdServer.Flag.Float64("master.garbageThreshold", 0.3, "threshold to vacuum and reclaim spaces")
masterOptions.metricsAddress = cmdServer.Flag.String("master.metrics.address", "", "Prometheus gateway address")
masterOptions.metricsIntervalSec = cmdServer.Flag.Int("master.metrics.intervalSeconds", 15, "Prometheus push interval in seconds")
masterOptions.raftResumeState = cmdServer.Flag.Bool("master.resumeState", true, "resume previous state on start master server")
masterOptions.raftHashicorp = cmdServer.Flag.Bool("master.raftHashicorp", false, "use hashicorp raft")
masterOptions.raftBootstrap = cmdServer.Flag.Bool("master.raftBootstrap", false, "Whether to bootstrap the Raft cluster")
masterOptions.heartbeatInterval = cmdServer.Flag.Duration("master.heartbeatInterval", 300*time.Millisecond, "heartbeat interval of master servers, and will be randomly multiplied by [1, 1.25)")
masterOptions.electionTimeout = cmdServer.Flag.Duration("master.electionTimeout", 10*time.Second, "election timeout of master servers")
masterOptions.telemetryUrl = cmdServer.Flag.String("master.telemetry.url", "https://telemetry.seaweedfs.com/api/collect", "telemetry server URL to send usage statistics")
masterOptions.telemetryEnabled = cmdServer.Flag.Bool("master.telemetry", true, "report anonymous cluster statistics to master.telemetry.url, use -master.telemetry=false to opt out")
filerOptions.filerGroup = cmdServer.Flag.String("filer.filerGroup", "", "share metadata with other filers in the same filerGroup")
filerOptions.collection = cmdServer.Flag.String("filer.collection", "", "all data will be stored in this collection")
filerOptions.port = cmdServer.Flag.Int("filer.port", 8888, "filer server http listen port")
filerOptions.portGrpc = cmdServer.Flag.Int("filer.port.grpc", 0, "filer server grpc listen port")
filerOptions.publicPort = cmdServer.Flag.Int("filer.port.public", 0, "filer server public http listen port")
filerOptions.disableHttp = cmdServer.Flag.Bool("filer.disableHttp", false, "disable filer HTTP requests, only gRPC operations are allowed")
filerOptions.allowedOrigins = cmdServer.Flag.String("filer.allowedOrigins", "*", "comma separated list of allowed origins")
filerOptions.defaultReplicaPlacement = cmdServer.Flag.String("filer.defaultReplicaPlacement", "", "default replication type. If not specified, use master setting.")
filerOptions.disableDirListing = cmdServer.Flag.Bool("filer.disableDirListing", false, "turn off directory listing")
filerOptions.maxMB = cmdServer.Flag.Int("filer.maxMB", 4, "split files larger than the limit")
filerOptions.dirListingLimit = cmdServer.Flag.Int("filer.dirListLimit", 1000, "limit sub dir listing size")
filerOptions.cipher = cmdServer.Flag.Bool("filer.encryptVolumeData", false, "encrypt data on volume servers")
filerOptions.saveToFilerLimit = cmdServer.Flag.Int("filer.saveToFilerLimit", 0, "Small files smaller than this limit can be cached in filer store.")
filerOptions.concurrentUploadLimitMB = cmdServer.Flag.Int("filer.concurrentUploadLimitMB", 0, "limit total concurrent upload size, 0 means unlimited")
filerOptions.concurrentFileUploadLimit = cmdServer.Flag.Int("filer.concurrentFileUploadLimit", 0, "limit number of concurrent file uploads, 0 means unlimited")
filerOptions.localSocket = cmdServer.Flag.String("filer.localSocket", "", "default to /tmp/seaweedfs-filer-<port>.sock")
filerOptions.showUIDirectoryDelete = cmdServer.Flag.Bool("filer.ui.deleteDir", true, "enable filer UI show delete directory button")
filerOptions.downloadMaxMBps = cmdServer.Flag.Int("filer.downloadMaxMBps", 0, "download max speed for each download request, in MB per second")
filerOptions.diskType = cmdServer.Flag.String("filer.disk", "", "[hdd|ssd|<tag>] hard drive or solid state drive or any tag")
filerOptions.exposeDirectoryData = cmdServer.Flag.Bool("filer.exposeDirectoryData", true, "expose directory data via filer. If false, filer UI will be inaccessible.")
filerOptions.tusBasePath = cmdServer.Flag.String("filer.tusBasePath", "/.tus", "TUS resumable upload endpoint base path (e.g., /.tus)")
filerOptions.tusMaxSizeMB = cmdServer.Flag.Int("filer.tusMaxSizeMB", 5*1024, "maximum TUS upload size in MB")
filerOptions.tusSessionExpiry = cmdServer.Flag.Duration("filer.tusSessionExpiry", 24*time.Hour, "incomplete TUS upload sessions are cleaned up after this duration, e.g. \"48h\", \"7h30m\"")
filerOptions.allowUntrustedRemoteEndpoints = cmdServer.Flag.Bool("filer.allowUntrustedRemoteEndpoints", false, allowUntrustedRemoteEndpointsUsage)
filerOptions.remoteCacheEvictThreshold = cmdServer.Flag.Float64("filer.remoteCacheEvictThreshold", 0.9, "evict remote-cached objects (oldest first) when any volume disk exceeds this usage fraction; 0 disables")
filerOptions.idleTimeout = cmdServer.Flag.Int("filer.idleTimeout", 10, "connection idle seconds")
serverOptions.v.port = cmdServer.Flag.Int("volume.port", 8080, "volume server http listen port")
serverOptions.v.portGrpc = cmdServer.Flag.Int("volume.port.grpc", 0, "volume server grpc listen port")
serverOptions.v.publicPort = cmdServer.Flag.Int("volume.port.public", 0, "volume server public port")
serverOptions.v.id = cmdServer.Flag.String("volume.id", "", "volume server id. If empty, default to ip:port")
serverOptions.v.indexType = cmdServer.Flag.String("volume.index", "memory", "Choose [memory|leveldb|leveldbMedium|leveldbLarge] mode for memory~performance balance.")
serverOptions.v.diskType = cmdServer.Flag.String("volume.disk", "", "[hdd|ssd|<tag>] hard drive or solid state drive or any tag")
serverOptions.v.tags = cmdServer.Flag.String("volume.tags", "", "comma-separated tag groups per data dir; each group uses ':' (e.g. fast:ssd,archive)")
serverOptions.v.fixJpgOrientation = cmdServer.Flag.Bool("volume.images.fix.orientation", false, "Adjust jpg orientation when uploading.")
serverOptions.v.readMode = cmdServer.Flag.String("volume.readMode", "proxy", "[local|proxy|redirect] how to deal with non-local volume: 'not found|read in remote node|redirect volume location'.")
serverOptions.v.compactionMBPerSecond = cmdServer.Flag.Int("volume.compactionMBps", 0, "limit compaction speed in mega bytes per second")
serverOptions.v.maintenanceMBPerSecond = cmdServer.Flag.Int("volume.maintenanceMBps", 0, "limit maintenance (replication / balance) IO rate in MB/s. Unset is 0, no limitation.")
serverOptions.v.maxParallelVacuumCommits = cmdServer.Flag.Int("volume.compactionMaxParallelCommits", 0, "limit how many vacuum commits run at once on this server; queued commits hold no volume lock. 0 means unlimited")
serverOptions.v.fileSizeLimitMB = cmdServer.Flag.Int("volume.fileSizeLimitMB", 256, "limit file size to avoid out of memory")
serverOptions.v.ldbTimeout = cmdServer.Flag.Int64("volume.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.")
serverOptions.v.concurrentUploadLimitMB = cmdServer.Flag.Int("volume.concurrentUploadLimitMB", 0, "limit total concurrent upload size, 0 means unlimited")
serverOptions.v.concurrentDownloadLimitMB = cmdServer.Flag.Int("volume.concurrentDownloadLimitMB", 0, "limit total concurrent download size, 0 means unlimited")
serverOptions.v.publicUrl = cmdServer.Flag.String("volume.publicUrl", "", "publicly accessible address")
serverOptions.v.preStopSeconds = cmdServer.Flag.Int("volume.preStopSeconds", 10, "number of seconds between stop send heartbeats and stop volume server")
serverOptions.v.pprof = cmdServer.Flag.Bool("volume.pprof", false, "enable pprof http handlers. precludes -memprofile and -cpuprofile")
serverOptions.v.idxFolder = cmdServer.Flag.String("volume.dir.idx", "", "directory to store .idx files")
serverOptions.v.inflightUploadDataTimeout = cmdServer.Flag.Duration("volume.inflightUploadDataTimeout", 60*time.Second, "inflight upload data wait timeout of volume servers")
serverOptions.v.inflightDownloadDataTimeout = cmdServer.Flag.Duration("volume.inflightDownloadDataTimeout", 60*time.Second, "inflight download data wait timeout of volume servers")
serverOptions.v.hasSlowRead = cmdServer.Flag.Bool("volume.hasSlowRead", true, "<experimental> if true, this prevents slow reads from blocking other requests, but large file read P99 latency will increase.")
serverOptions.v.readBufferSizeMB = cmdServer.Flag.Int("volume.readBufferSizeMB", 4, "<experimental> larger values can optimize query performance but will increase some memory usage,Use with hasSlowRead normally")
serverOptions.v.allowUntrustedRemoteEndpoints = cmdServer.Flag.Bool("volume.allowUntrustedRemoteEndpoints", false, "if true, FetchAndWriteNeedle accepts arbitrary remote S3 endpoints including loopback / link-local hosts. Default rejects internal / metadata endpoints.")
serverOptions.v.fsyncMinBatchBytes = cmdServer.Flag.Int64("volume.fsyncMinBatchBytes", 1*1024*1024, "minimum bytes an fsync batch gathers before a drained queue flushes without waiting")
serverOptions.v.fsyncMinBatchRequests = cmdServer.Flag.Int("volume.fsyncMinBatchRequests", 128, "minimum requests an fsync batch gathers before a drained queue flushes without waiting")
serverOptions.v.fsyncBatchWait = cmdServer.Flag.Duration("volume.fsyncBatchWait", 500*time.Microsecond, "how long a below-minimum fsync batch waits for more requests once the queue drains; 0 flushes immediately")
serverOptions.v.setDiskIOProbeDefaults()
s3Options.port = cmdServer.Flag.Int("s3.port", 8333, "s3 server http listen port")
s3Options.portHttps = cmdServer.Flag.Int("s3.port.https", 0, "s3 server https listen port")
s3Options.portGrpc = cmdServer.Flag.Int("s3.port.grpc", 0, "s3 server grpc listen port")
s3Options.portIceberg = cmdServer.Flag.Int("s3.port.iceberg", 8181, "Iceberg REST Catalog server listen port (0 to disable)")
s3Options.portLance = cmdServer.Flag.Int("s3.port.lance", 9101, "Lance Namespace server listen port (0 to disable)")
s3Options.icebergCredentialRole = cmdServer.Flag.String("s3.iceberg.credentialRole", "", "IAM role ARN the Iceberg catalog assumes to vend table-scoped credentials (empty disables vending)")
s3Options.icebergCredentialDuration = cmdServer.Flag.Int("s3.iceberg.credentialDurationSeconds", 3600, "lifetime of credentials vended by the Iceberg catalog")
s3Options.domainName = cmdServer.Flag.String("s3.domainName", "", "suffix of the host name in comma separated list, {bucket}.{domainName}")
s3Options.allowedOrigins = cmdServer.Flag.String("s3.allowedOrigins", "*", "comma separated list of allowed origins")
s3Options.dataCenter = cmdServer.Flag.String("s3.dataCenter", "", "prefer to read and write to volumes in this data center")
s3Options.tlsPrivateKey = cmdServer.Flag.String("s3.key.file", "", "path to the TLS private key file")
s3Options.tlsCertificate = cmdServer.Flag.String("s3.cert.file", "", "path to the TLS certificate file")
s3Options.tlsCACertificate = cmdServer.Flag.String("s3.cacert.file", "", "path to the TLS CA certificate file")
s3Options.tlsVerifyClientCert = cmdServer.Flag.Bool("s3.tlsVerifyClientCert", false, "whether to verify the client's certificate")
s3Options.config = cmdServer.Flag.String("s3.config", "", "path to the config file")
s3Options.iamConfig = cmdServer.Flag.String("s3.iam.config", "", "path to the advanced IAM config file for S3. Overrides -iam.config if both are provided.")
s3Options.auditLogConfig = cmdServer.Flag.String("s3.auditLogConfig", "", "path to the audit log config file")
cmdServer.Flag.Bool("s3.allowEmptyFolder", true, "deprecated, ignored. Empty folder cleanup is now automatic.")
s3Options.allowDeleteBucketNotEmpty = cmdServer.Flag.Bool("s3.allowDeleteBucketNotEmpty", false, "allow recursive deleting all entries along with bucket")
s3Options.autoCreateBucket = cmdServer.Flag.Bool("s3.autoCreateBucket", true, "create the bucket on upload if it does not exist, for admin identities only")
s3Options.localSocket = cmdServer.Flag.String("s3.localSocket", "", "default to /tmp/seaweedfs-s3-<port>.sock")
s3Options.bindIp = cmdServer.Flag.String("s3.ip.bind", "", "ip address to bind to. If empty, default to same as -ip.bind option.")
s3Options.idleTimeout = cmdServer.Flag.Int("s3.idleTimeout", 120, "connection idle seconds")
s3Options.concurrentUploadLimitMB = cmdServer.Flag.Int("s3.concurrentUploadLimitMB", 0, "limit total concurrent upload size for S3, 0 means unlimited")
s3Options.concurrentFileUploadLimit = cmdServer.Flag.Int("s3.concurrentFileUploadLimit", 0, "limit number of concurrent file uploads for S3, 0 means unlimited")
s3Options.enableIam = cmdServer.Flag.Bool("s3.iam", true, "enable embedded IAM API on the same S3 port")
s3Options.iamReadOnly = cmdServer.Flag.Bool("s3.iam.readOnly", true, "disable IAM write operations on this server")
s3Options.cipher = cmdServer.Flag.Bool("s3.encryptVolumeData", false, "encrypt data on volume servers for S3 uploads")
s3Options.externalUrl = cmdServer.Flag.String("s3.externalUrl", "", "the external URL clients use to connect (e.g. https://api.example.com:9000). Advertised to Iceberg and Lance clients, and tried first when verifying S3 signatures behind a reverse proxy. Falls back to S3_EXTERNAL_URL env var.")
s3Options.allowBackslashInKey = cmdServer.Flag.Bool("s3.allowBackslashInKey", false, allowBackslashInKeyUsage)
s3Options.defaultFileMode = cmdServer.Flag.String("s3.defaultFileMode", "", "default file mode for S3 uploaded objects, e.g. 0660, 0644, 0666")
s3Options.cacheSizeMB = cmdServer.Flag.Int64("s3.cacheCapacityMB", 0, "in-memory chunk cache capacity in MB for S3 GETs shared across requests (0 disables)")
s3Options.readerCacheSizeMB = cmdServer.Flag.Int64("s3.readerCacheSizeMB", 0, "memory budget in MiB for downloaded and in-flight reader buffers across all S3 GETs (0 means unlimited)")
s3Options.allowUntrustedRemoteEndpoints = cmdServer.Flag.Bool("s3.allowUntrustedRemoteEndpoints", false, allowUntrustedRemoteEndpointsUsage)
s3Options.iamSelfServiceRequiresGrant = cmdServer.Flag.Bool("s3.iam.selfServiceRequiresGrant", false, iamSelfServiceRequiresGrantUsage)
sftpOptions.port = cmdServer.Flag.Int("sftp.port", 2022, "SFTP server listen port")
sftpOptions.idleTimeout = cmdServer.Flag.Int("sftp.idleTimeout", 10, "connection idle seconds")
sftpOptions.sshPrivateKey = cmdServer.Flag.String("sftp.sshPrivateKey", "", "path to the SSH private key file for host authentication")
sftpOptions.hostKeysFolder = cmdServer.Flag.String("sftp.hostKeysFolder", "", "path to folder containing SSH private key files for host authentication")
sftpOptions.authMethods = cmdServer.Flag.String("sftp.authMethods", "password,publickey", "comma-separated list of allowed auth methods: password, publickey, certificate")
sftpOptions.maxAuthTries = cmdServer.Flag.Int("sftp.maxAuthTries", 6, "maximum number of authentication attempts per connection")
sftpOptions.bannerMessage = cmdServer.Flag.String("sftp.bannerMessage", "SeaweedFS SFTP Server - Unauthorized access is prohibited", "message displayed before authentication")
sftpOptions.loginGraceTime = cmdServer.Flag.Duration("sftp.loginGraceTime", 2*time.Minute, "timeout for authentication")
sftpOptions.clientAliveInterval = cmdServer.Flag.Duration("sftp.clientAliveInterval", 5*time.Second, "interval for sending keep-alive messages")
sftpOptions.clientAliveCountMax = cmdServer.Flag.Int("sftp.clientAliveCountMax", 3, "maximum number of missed keep-alive messages before disconnecting")
sftpOptions.userStoreFile = cmdServer.Flag.String("sftp.userStoreFile", "", "path to JSON file containing user credentials and permissions")
sftpOptions.trustedUserCAKeysFile = cmdServer.Flag.String("sftp.trustedUserCAKeysFile", "", "path to a file with trusted user CA public keys (OpenSSH authorized_keys format); required when 'certificate' is in -sftp.authMethods")
sftpOptions.dataCenter = cmdServer.Flag.String("sftp.dataCenter", "", "prefer to read and write to volumes in this data center")
sftpOptions.localSocket = cmdServer.Flag.String("sftp.localSocket", "", "default to /tmp/seaweedfs-sftp-<port>.sock")
iamOptions.port = cmdServer.Flag.Int("iam.port", 8111, "iam server http listen port")
iamOptions.idleTimeout = cmdServer.Flag.Int("iam.idleTimeout", 10, "connection idle seconds")
iamOptions.selfServiceRequiresGrant = s3Options.iamSelfServiceRequiresGrant
webdavOptions.port = cmdServer.Flag.Int("webdav.port", 7333, "webdav server http listen port")
webdavOptions.collection = cmdServer.Flag.String("webdav.collection", "", "collection to create the files")
webdavOptions.replication = cmdServer.Flag.String("webdav.replication", "", "replication to create the files")
webdavOptions.disk = cmdServer.Flag.String("webdav.disk", "", "[hdd|ssd|<tag>] hard drive or solid state drive or any tag")
webdavOptions.tlsPrivateKey = cmdServer.Flag.String("webdav.key.file", "", "path to the TLS private key file")
webdavOptions.tlsCertificate = cmdServer.Flag.String("webdav.cert.file", "", "path to the TLS certificate file")
webdavOptions.cacheDir = cmdServer.Flag.String("webdav.cacheDir", os.TempDir(), "local cache directory for file chunks")
webdavOptions.cacheSizeMB = cmdServer.Flag.Int64("webdav.cacheCapacityMB", 0, "local cache capacity in MB")
webdavOptions.maxMB = cmdServer.Flag.Int("webdav.maxMB", 4, "split files larger than the limit")
webdavOptions.filerRootPath = cmdServer.Flag.String("webdav.filer.path", "/", "use this remote path from filer server")
webdavOptions.idleTimeout = cmdServer.Flag.Int("webdav.idleTimeout", 10, "connection idle seconds")
mqBrokerOptions.port = cmdServer.Flag.Int("mq.broker.port", 17777, "message queue broker gRPC listen port")
mqBrokerOptions.logFlushInterval = cmdServer.Flag.Int("mq.broker.logFlushInterval", 5, "log buffer flush interval in seconds")
mqAgentServerOptions.brokersString = cmdServer.Flag.String("mq.agent.brokers", "localhost:17777", "comma-separated message queue brokers")
mqAgentServerOptions.port = cmdServer.Flag.Int("mq.agent.port", 16777, "message queue agent gRPC listen port")
}
// isServerFlagPassed checks if a specific flag was passed on the command line
func isServerFlagPassed(name string) bool {
found := false
cmdServer.Flag.Visit(func(f *flag.Flag) {
if f.Name == name {
found = true
}
})
return found
}
func runServer(cmd *Command, args []string) bool {
if *serverOptions.debug {
go http.ListenAndServe(fmt.Sprintf(":%d", *serverOptions.debugPort), nil)
}
util.LoadSecurityConfiguration()
util.LoadConfiguration("master", false)
util.LoadConfiguration("volume", false)
serverOptions.v.applyDiskIOProbeConfig()
*serverOptions.cpuprofile = util.ResolvePath(*serverOptions.cpuprofile)
*serverOptions.memprofile = util.ResolvePath(*serverOptions.memprofile)
*serverIamConfig = util.ResolvePath(*serverIamConfig)
*masterOptions.metaFolder = util.ResolvePath(*masterOptions.metaFolder)
s3Options.resolvePaths()
webdavOptions.resolvePaths()
sftpOptions.resolvePaths()
grace.SetupProfiling(*serverOptions.cpuprofile, *serverOptions.memprofile)
if *isStartingS3 {
*isStartingFiler = true
}
if *isStartingSftp {
*isStartingFiler = true
}
if *isStartingIam {
*isStartingFiler = true
}
if *isStartingWebDav {
*isStartingFiler = true
}
if *isStartingMqBroker {
*isStartingFiler = true
}
if *isStartingMqAgent {
*isStartingMqBroker = true
*isStartingFiler = true
}
var actualPeersForComponents string
var embeddedMaster atomic.Pointer[weed_server.MasterServer]
var peerList []pb.ServerAddress
if *isStartingMasterServer {
// If we are starting a master, validate and complete the peer list
_, peerList = checkPeers(*serverIp, *masterOptions.port, *masterOptions.portGrpc, *masterOptions.peers)
actualPeersForComponents = strings.Join(pb.ToAddressStrings(peerList), ",")
if len(peerList) == 1 && *isStartingVolumeServer {
// A single-master weed server has one guaranteed reporter: its
// embedded volume server, whose first heartbeat already carries the
// full inventory. Its clean report collapses the post-mint
// allocation hold to a confirm pulse instead of running the whole
// warmup window on every fresh start. Other volume servers stay
// admitted, so a late veteran still lands its prior-state evidence
// inside that pulse.
masterOptions.expectedReporterAddr = util.JoinHostPort(*serverIp, *serverOptions.v.port)
}
} else if *masterOptions.peers != "" {
if isSingleMasterMode(*masterOptions.peers) {
glog.Fatalf("'-master.peers=none' is only valid when starting a master server, but master is not starting.")
}
// If not starting a master, just use the provided peers
actualPeersForComponents = *masterOptions.peers
}
// The loopback companion listener is only kept while -ip.bind is unset and
// the bind address falls back to -ip. Capture that here: serverBindIp is
// repointed at serverIp below, after which the flag state is no longer
// distinguishable.
ipBindUnspecified := *serverBindIp == ""
if *serverBindIp == "" {
serverBindIp = serverIp
}
// Bind outbound connections to the same address up front so every
// component started below inherits it, before any of them dials.
util.SetOutboundLocalIP(*serverBindIp)
if *serverMetricsHttpIp == "" {
*serverMetricsHttpIp = *serverBindIp
}
// ip address
masterOptions.ip = serverIp
masterOptions.ipBind = serverBindIp
masterOptions.listenOnLocalhost = &ipBindUnspecified
// Use actualPeersForComponents for volume/filer, not masterOptions.peers which might be "none"
filerOptions.masters = pb.ServerAddresses(actualPeersForComponents).ToServiceDiscovery()
filerOptions.ip = serverIp
filerOptions.bindIp = serverBindIp
filerOptions.listenOnLocalhost = &ipBindUnspecified
s3Options.ip = serverIp
if *s3Options.bindIp == "" {
s3Options.listenOnLocalhost = &ipBindUnspecified
s3Options.bindIp = serverBindIp
}
if sftpOptions.bindIp == nil || *sftpOptions.bindIp == "" {
sftpOptions.listenOnLocalhost = &ipBindUnspecified
sftpOptions.bindIp = serverBindIp
}
iamOptions.ip = serverBindIp
iamOptions.listenOnLocalhost = &ipBindUnspecified
iamOptions.masters = &actualPeersForComponents
webdavOptions.ipBind = serverBindIp
serverOptions.v.ip = serverIp
serverOptions.v.bindIp = serverBindIp
serverOptions.v.listenOnLocalhost = &ipBindUnspecified
serverOptions.v.masters = pb.ServerAddresses(actualPeersForComponents).ToAddresses()
serverOptions.v.idleConnectionTimeout = serverTimeout
serverOptions.v.dataCenter = serverDataCenter
serverOptions.v.rack = serverRack
mqBrokerOptions.ip = serverIp
mqBrokerOptions.ipBind = serverBindIp
mqBrokerOptions.listenOnLocalhost = &ipBindUnspecified
mqBrokerOptions.masters = filerOptions.masters.GetInstancesAsMap()
mqBrokerOptions.filerGroup = filerOptions.filerGroup
mqAgentServerOptions.ip = serverIp
mqAgentServerOptions.ipBind = serverBindIp
mqAgentServerOptions.listenOnLocalhost = &ipBindUnspecified
if isServerFlagPassed("mq.agent.brokers") {
mqAgentServerOptions.brokers = pb.ServerAddresses(*mqAgentServerOptions.brokersString).ToAddresses()
} else {
// Default to the embedded broker's bind address: with an explicit
// -ip.bind the broker does not listen on localhost.
brokerHost := *serverBindIp
if brokerHost == "0.0.0.0" || brokerHost == "::" || brokerHost == "[::]" {
brokerHost = "localhost"
}
mqAgentServerOptions.brokers = []pb.ServerAddress{pb.NewServerAddress(brokerHost, *mqBrokerOptions.port, 0)}
}
// serverOptions.v.pulseSeconds = pulseSeconds
// masterOptions.pulseSeconds = pulseSeconds
masterOptions.whiteList = serverWhiteListOption
// The master's /submit buffers the upload before handing it to the volume
// server started here, so it must not reject what that volume server accepts.
masterOptions.fileSizeLimitMB = serverOptions.v.fileSizeLimitMB
filerOptions.dataCenter = serverDataCenter
filerOptions.rack = serverRack
mqBrokerOptions.dataCenter = serverDataCenter
mqBrokerOptions.rack = serverRack
*filerOptions.disableHttp = *filerOptions.disableHttp || *serverDisableHttp
masterOptions.disableHttp = serverDisableHttp
filerAddress := string(pb.NewServerAddress(*serverIp, *filerOptions.port, *filerOptions.portGrpc))
s3Options.filer = &filerAddress
sftpOptions.filer = &filerAddress
iamOptions.filer = &filerAddress
webdavOptions.filer = &filerAddress
mqBrokerOptions.filerGroup = filerOptions.filerGroup
go stats_collect.StartMetricsServer(*serverMetricsHttpIp, *serverMetricsHttpPort)
*volumeDataFolders = util.ResolveCommaSeparatedPaths(*volumeDataFolders)
folders := strings.Split(*volumeDataFolders, ",")
if *masterOptions.volumeSizeLimitMB > util.MaxVolumeSizeLimitMB {
glog.Fatalf("master.volumeSizeLimitMB should not exceed %d", util.MaxVolumeSizeLimitMB)
}
if *masterOptions.metaFolder == "" {
*masterOptions.metaFolder = folders[0]
}
if err := util.TestFolderWritable(*masterOptions.metaFolder); err != nil {
glog.Fatalf("Check Meta Folder (-mdir=\"%s\") Writable: %s", *masterOptions.metaFolder, err)
}
filerOptions.defaultLevelDbDirectory = masterOptions.metaFolder
// Register Unix socket paths for gRPC services running in this process
// so local inter-service communication uses Unix sockets instead of TCP.
// Resolve gRPC ports early (same calculation each service does internally).
for _, svc := range []struct {
starting *bool
portGrpc *int
port *int
name string
}{
{isStartingMasterServer, masterOptions.portGrpc, masterOptions.port, "master"},
{isStartingVolumeServer, serverOptions.v.portGrpc, serverOptions.v.port, "volume"},
{isStartingFiler, filerOptions.portGrpc, filerOptions.port, "filer"},
{isStartingS3, s3Options.portGrpc, s3Options.port, "s3"},
} {
if *svc.starting {
if *svc.portGrpc == 0 {
*svc.portGrpc = 10000 + *svc.port
}
pb.RegisterLocalGrpcSocket(*serverIp, *svc.portGrpc, fmt.Sprintf("/tmp/seaweedfs-%s-grpc-%d.sock", svc.name, *svc.portGrpc))
}
}
serverWhiteList := util.StringSplit(*serverWhiteListOption, ",")
if *isStartingFiler {
if *filerOptions.diskType == "" && *serverOptions.v.diskType != "" {
filerOptions.diskType = serverOptions.v.diskType
}
if len(peerList) == 1 && *isStartingMasterServer {
// The embedded filer shares the master's process: when the master
// mints a fresh TopologyId its persisted state was empty, so the
// empty peer scan only has to survive a reconnecting peer's
// keepalive retries, not the full membership confirm. Prior state
// keeps the full window.
filerOptions.clusterFreshHint = func() bool {
ms := embeddedMaster.Load()
return ms != nil && ms.Topo.FreshMinted()
}
}
go func() {
time.Sleep(1 * time.Second)
filerOptions.startFiler()
}()
}
if *isStartingS3 {
// Handle IAM config: -s3.iam.config takes precedence over -iam.config
if *s3Options.iamConfig == "" {
*s3Options.iamConfig = *serverIamConfig
} else if *serverIamConfig != "" && *s3Options.iamConfig != *serverIamConfig {
glog.V(0).Infof("both -s3.iam.config(%s) and -iam.config(%s) provided; using -s3.iam.config", *s3Options.iamConfig, *serverIamConfig)
}
// Share the S3 static identity config file with the filer
filerOptions.s3ConfigFile = s3Options.config
go func() {
time.Sleep(2 * time.Second)
s3Options.localFilerSocket = filerOptions.localSocket
s3Options.startS3Server()
}()
}
if *isStartingSftp {
go func() {
time.Sleep(2 * time.Second)
sftpOptions.startSftpServer()
}()
}
if *isStartingIam {
go func() {
time.Sleep(2 * time.Second)
iamOptions.startIamServer()
}()
}
if *isStartingWebDav {
go func() {
time.Sleep(2 * time.Second)
webdavOptions.startWebDav()
}()
}
if *isStartingMqBroker {
go func() {
time.Sleep(2 * time.Second)
mqBrokerOptions.startQueueServer()
}()
}
if *isStartingMqAgent {
go func() {
time.Sleep(2 * time.Second)
mqAgentServerOptions.startQueueAgent()
}()
}
// start volume server
if *isStartingVolumeServer {
minFreeSpaces := util.MustParseMinFreeSpace(*volumeMinFreeSpace, *volumeMinFreeSpacePercent)
go serverOptions.v.startVolumeServer(*volumeDataFolders, *volumeMaxDataVolumeCounts, *serverWhiteListOption, minFreeSpaces)
}
if *isStartingMasterServer {
go startMaster(masterOptions, serverWhiteList, &embeddedMaster)
}
select {}
}
// idleTimeoutSeconds converts a seconds flag to the listener activity timeout,
// keeping the 10s default when the option was never set. The clamp keeps the
// multiplication from overflowing time.Duration; non-positive values disable
// the timeout, as before.
func idleTimeoutSeconds(seconds *int) time.Duration {
if seconds == nil {
return 10 * time.Second
}
if *seconds <= 0 {
return 0
}
return time.Duration(min(int64(*seconds), math.MaxInt64/int64(time.Second))) * time.Second
}
func newHttpServer(h http.Handler, tlsConfig *tls.Config) *http.Server {
s := &http.Server{
Handler: h,
}
if tlsConfig != nil {
s.TLSConfig = tlsConfig.Clone()
}
return s
}
Reference in new issue View Git Blame Copy Permalink