mirror of
https://github.com/hanzoai/s3.git
synced 2026-11-03 03:07:41 +00:00
zooqueen
dd74d3e6d6
hanzoai/s3 was an AGPL-3.0 MinIO fork whose network-copyleft blocks embedding
and reselling as a managed service. Replace it wholesale with a fork of
SeaweedFS (github.com/seaweedfs/seaweedfs, Apache-2.0, Go) which ships a
built-in S3 gateway + Filer.
- License: AGPL-3.0 -> Apache-2.0 (upstream LICENSE kept verbatim, incl.
"Copyright 2025 Chris Lu"); NOTICE added crediting SeaweedFS. No AGPL/BUSL/
SSPL/GPL or enterprise ("ee") files remain.
- Module: go.mod path set to github.com/hanzoai/s3 with a self-replace
(replace github.com/seaweedfs/seaweedfs => ./) so the ~1800 internal Go files
keep their upstream import paths with zero churn. Internal identifiers and
import paths are intentionally NOT renamed.
- CLI rebrand: the binary is built as `s3` (was `weed`); root usage, per-command
help/usage/flag strings, shell autocomplete program name, and the product name
in CLI output now read "Hanzo S3" / `s3`. Users never see `weed`.
(e.g. `s3 server`, `s3 filer`, `s3 master`, `s3 volume`, `s3 s3`, `s3 mount`.)
- Build: `make` (or `go build -o s3 ./weed`) builds the `s3` binary green;
`cmd/weed-db` and `cmd/weed-sql` also build; `go vet ./weed/command/` clean.
- README/LLM.md rewritten for Hanzo S3; AGENTS.md/CLAUDE.md symlinks preserved.
- Dropped a 33MB prebuilt Mach-O `stress` binary upstream had committed; `.gitignore`
updated to ignore the `s3` build output.
116 lines
3.7 KiB
Go
116 lines
3.7 KiB
Go
package topology
|
|
|
|
import (
|
|
"math/rand/v2"
|
|
"time"
|
|
|
|
"github.com/seaweedfs/seaweedfs/weed/stats"
|
|
"github.com/seaweedfs/seaweedfs/weed/storage/erasure_coding"
|
|
"github.com/seaweedfs/seaweedfs/weed/storage/types"
|
|
"google.golang.org/grpc"
|
|
|
|
"github.com/seaweedfs/seaweedfs/weed/glog"
|
|
"github.com/seaweedfs/seaweedfs/weed/storage"
|
|
)
|
|
|
|
func (t *Topology) StartRefreshWritableVolumes(grpcDialOption grpc.DialOption, garbageThreshold float64, concurrentVacuumLimitPerVolumeServer int, growThreshold float64, preallocate int64) {
|
|
go func() {
|
|
for {
|
|
if t.IsLeader() {
|
|
freshThreshHold := time.Now().Unix() - 3*t.pulse //3 times of sleep interval
|
|
t.CollectDeadNodeAndFullVolumes(freshThreshHold, t.volumeSizeLimit, growThreshold)
|
|
}
|
|
time.Sleep(time.Duration(float32(t.pulse*1e3)*(1+rand.Float32())) * time.Millisecond)
|
|
}
|
|
}()
|
|
go func(garbageThreshold float64) {
|
|
for {
|
|
if t.IsLeader() {
|
|
// Safety net: if vacuum was disabled by the plugin monitor but the
|
|
// admin server is no longer connected, automatically re-enable.
|
|
// This handles the case where the admin server crashes without
|
|
// cleanup. Does NOT override an operator's intentional disable.
|
|
if t.IsVacuumDisabledByPlugin() && t.adminServerConnectedFunc != nil && !t.adminServerConnectedFunc() {
|
|
glog.V(0).Infof("Admin server disconnected while vacuum was disabled by plugin, clearing plugin disable")
|
|
t.EnableVacuumByPlugin()
|
|
}
|
|
if !t.IsVacuumDisabled() {
|
|
t.Vacuum(grpcDialOption, garbageThreshold, concurrentVacuumLimitPerVolumeServer, 0, "", preallocate, true)
|
|
}
|
|
} else {
|
|
stats.MasterReplicaPlacementMismatch.Reset()
|
|
}
|
|
time.Sleep(14*time.Minute + time.Duration(120*rand.Float32())*time.Second)
|
|
}
|
|
}(garbageThreshold)
|
|
go func() {
|
|
for {
|
|
select {
|
|
case fv := <-t.chanFullVolumes:
|
|
t.SetVolumeCapacityFull(fv)
|
|
case cv := <-t.chanCrowdedVolumes:
|
|
t.SetVolumeCrowded(cv)
|
|
}
|
|
}
|
|
}()
|
|
}
|
|
func (t *Topology) SetVolumeCapacityFull(volumeInfo storage.VolumeInfo) bool {
|
|
diskType := types.ToDiskType(volumeInfo.DiskType)
|
|
vl := t.GetVolumeLayout(volumeInfo.Collection, volumeInfo.ReplicaPlacement, volumeInfo.Ttl, diskType)
|
|
if !vl.SetVolumeCapacityFull(volumeInfo.Id) {
|
|
return false
|
|
}
|
|
|
|
vl.accessLock.RLock()
|
|
defer vl.accessLock.RUnlock()
|
|
|
|
vidLocations, found := vl.vid2location[volumeInfo.Id]
|
|
if !found {
|
|
return false
|
|
}
|
|
|
|
for _, dn := range vidLocations.list {
|
|
if !volumeInfo.ReadOnly {
|
|
|
|
disk := dn.getOrCreateDisk(volumeInfo.DiskType)
|
|
disk.UpAdjustDiskUsageDelta(types.ToDiskType(volumeInfo.DiskType), &DiskUsageCounts{
|
|
activeVolumeCount: -1,
|
|
})
|
|
|
|
}
|
|
}
|
|
return true
|
|
}
|
|
|
|
func (t *Topology) SetVolumeCrowded(volumeInfo storage.VolumeInfo) {
|
|
diskType := types.ToDiskType(volumeInfo.DiskType)
|
|
vl := t.GetVolumeLayout(volumeInfo.Collection, volumeInfo.ReplicaPlacement, volumeInfo.Ttl, diskType)
|
|
vl.SetVolumeCrowded(volumeInfo.Id)
|
|
}
|
|
|
|
func (t *Topology) UnRegisterDataNode(dn *DataNode) {
|
|
dn.IsTerminating = true
|
|
for _, v := range dn.GetVolumes() {
|
|
glog.V(0).Infoln("Removing Volume", v.Id, "from the dead volume server", dn.Id())
|
|
diskType := types.ToDiskType(v.DiskType)
|
|
vl := t.GetVolumeLayout(v.Collection, v.ReplicaPlacement, v.Ttl, diskType)
|
|
vl.SetVolumeUnavailable(dn, v.Id)
|
|
}
|
|
|
|
// unregister ec shards when volume server disconnected
|
|
for _, s := range dn.GetEcShards() {
|
|
t.UnRegisterEcShards(s, dn)
|
|
}
|
|
|
|
negativeUsages := dn.GetDiskUsages().negative()
|
|
for dt, du := range negativeUsages.usages {
|
|
dn.UpAdjustDiskUsageDelta(dt, du)
|
|
}
|
|
dn.DeltaUpdateVolumes([]storage.VolumeInfo{}, dn.GetVolumes())
|
|
dn.DeltaUpdateEcShards([]*erasure_coding.EcVolumeInfo{}, dn.GetEcShards())
|
|
t.unregisterDataNodeAddress(dn.ServerAddress(), dn)
|
|
if dn.Parent() != nil {
|
|
dn.Parent().UnlinkChildNode(dn.Id())
|
|
}
|
|
}
|