mirror of
https://github.com/seaweedfs/seaweedfs.git
synced 2026-10-21 13:34:26 +02:00
Chris Lu
2874bd05fb
* pb: ResumeVolumeGrowth rpc for operator-driven growth resume (#11763) * master: freeze volume growth on topology id mismatch (#11763) * shell: volume.growth.resume command (#11763) * master: replicate the growth freeze and gate it on veteran heartbeats (#11763) * master: block allocation while the topology id is being minted DeferAllocation only armed after the topology id persisted, so growth requests arriving during init ran ahead of the warmup window. Gate NextVolumeId on an init-pending flag spanning the whole mint. * master: stash reported topology ids until the local id is minted (#11763) * drain stashed topology id reports when the id arrives outside the mint path * persist the growth-freeze flag in the mirror file and raft snapshots * log hashicorp raft marshal and apply errors in applyGrowthFreeze
72 lines
2.2 KiB
Go
72 lines
2.2 KiB
Go
package weed_server
|
|
|
|
import (
|
|
"encoding/json"
|
|
"time"
|
|
|
|
"github.com/google/uuid"
|
|
"github.com/seaweedfs/seaweedfs/weed/glog"
|
|
"github.com/seaweedfs/seaweedfs/weed/topology"
|
|
)
|
|
|
|
// recoverTopologyIdFromState restores the TopologyId from serialized FSM
|
|
// state bytes (JSON-encoded MaxVolumeIdCommand). Both raft implementations
|
|
// call this after reading their snapshot in their own format.
|
|
func recoverTopologyIdFromState(fsmState []byte, topo *topology.Topology) {
|
|
if topo.GetTopologyId() != "" {
|
|
return
|
|
}
|
|
var cmd topology.MaxVolumeIdCommand
|
|
if err := json.Unmarshal(fsmState, &cmd); err != nil {
|
|
return
|
|
}
|
|
if cmd.TopologyId != "" {
|
|
topo.SetTopologyId(cmd.TopologyId)
|
|
glog.V(0).Infof("Recovered TopologyId from snapshot: %s", cmd.TopologyId)
|
|
}
|
|
}
|
|
|
|
// EnsureTopologyId ensures that a TopologyId is generated and persisted if it's currently missing.
|
|
// It uses the provided checkLeaderFn to verify leadership and persistFn to save the new ID.
|
|
func EnsureTopologyId(topo *topology.Topology, checkLeaderFn func() bool, persistFn func(string) error) {
|
|
if topo.GetTopologyId() != "" {
|
|
// The id arrived through raft replay or the mirror file; reports
|
|
// stashed while it was unknown can now be checked.
|
|
topo.SetTopologyInitPending(false)
|
|
return
|
|
}
|
|
|
|
topo.SetTopologyInitPending(true)
|
|
defer topo.SetTopologyInitPending(false)
|
|
|
|
topologyId := uuid.New().String()
|
|
for {
|
|
if !checkLeaderFn() {
|
|
glog.V(0).Infof("lost leadership while saving topologyId")
|
|
return
|
|
}
|
|
|
|
// Another concurrent operation may have set the ID between generation and now.
|
|
if latestId := topo.GetTopologyId(); latestId != "" {
|
|
glog.V(1).Infof("topologyId was set concurrently to %s, aborting generation", latestId)
|
|
return
|
|
}
|
|
|
|
if err := persistFn(topologyId); err != nil {
|
|
glog.Errorf("failed to save topologyId, will retry: %v", err)
|
|
time.Sleep(time.Second)
|
|
continue
|
|
}
|
|
|
|
// Verify that the topology ID was actually applied as expected.
|
|
appliedId := topo.GetTopologyId()
|
|
if appliedId != "" && appliedId != topologyId {
|
|
glog.V(0).Infof("TopologyId generation race: expected %s, but current TopologyId is %s", topologyId, appliedId)
|
|
} else {
|
|
glog.V(0).Infof("TopologyId generated: %s", topologyId)
|
|
}
|
|
topo.DeferAllocation(topo.WarmupDuration())
|
|
break
|
|
}
|
|
}
|