open-nomad/nomad/serf.go

199 lines
4.9 KiB
Go
Raw Normal View History

2015-06-03 10:58:00 +00:00
package nomad
import (
"sync/atomic"
"github.com/hashicorp/serf/serf"
)
2015-06-03 10:58:00 +00:00
2015-06-04 10:42:56 +00:00
const (
// StatusReap is used to update the status of a node if we
// are handling a EventMemberReap
StatusReap = serf.MemberStatus(-1)
)
2015-06-03 10:58:00 +00:00
// serfEventHandler is used to handle events from the serf cluster
func (s *Server) serfEventHandler() {
for {
select {
case e := <-s.eventCh:
switch e.EventType() {
case serf.EventMemberJoin:
s.nodeJoin(e.(serf.MemberEvent))
2015-06-04 10:42:56 +00:00
s.localMemberEvent(e.(serf.MemberEvent))
2015-06-03 10:58:00 +00:00
case serf.EventMemberLeave, serf.EventMemberFailed:
s.nodeFailed(e.(serf.MemberEvent))
2015-06-04 10:42:56 +00:00
s.localMemberEvent(e.(serf.MemberEvent))
2015-06-03 10:58:00 +00:00
case serf.EventMemberUpdate, serf.EventMemberReap,
serf.EventUser, serf.EventQuery: // Ignore
default:
s.logger.Printf("[WARN] nomad: unhandled serf event: %#v", e)
}
case <-s.shutdownCh:
return
}
}
}
// nodeJoin is used to handle join events on the serf cluster
func (s *Server) nodeJoin(me serf.MemberEvent) {
2015-06-04 10:33:12 +00:00
for _, m := range me.Members {
ok, parts := isNomadServer(m)
if !ok {
s.logger.Printf("[WARN] nomad: non-server in gossip pool: %s", m.Name)
continue
}
s.logger.Printf("[INFO] nomad: adding server %s", parts)
2015-06-03 10:58:00 +00:00
2015-06-04 10:33:12 +00:00
// Check if this server is known
found := false
s.peerLock.Lock()
existing := s.peers[parts.Region]
for idx, e := range existing {
if e.Name == parts.Name {
existing[idx] = parts
found = true
break
}
}
2015-06-03 10:58:00 +00:00
2015-06-04 10:33:12 +00:00
// Add ot the list if not known
if !found {
s.peers[parts.Region] = append(existing, parts)
}
2015-06-07 18:37:59 +00:00
// Check if a local peer
if parts.Region == s.config.Region {
s.localPeers[parts.Addr.String()] = parts
}
2015-06-04 10:33:12 +00:00
s.peerLock.Unlock()
2015-06-03 10:58:00 +00:00
2015-06-04 10:33:12 +00:00
// If we still expecting to bootstrap, may need to handle this
if atomic.LoadInt32(&s.config.BootstrapExpect) != 0 {
2015-06-04 10:33:12 +00:00
s.maybeBootstrap()
}
}
2015-06-03 10:58:00 +00:00
}
2016-06-16 00:23:02 +00:00
// maybeBootsrap is used to handle bootstrapping when a new server joins
2015-06-03 10:58:00 +00:00
func (s *Server) maybeBootstrap() {
2015-06-05 22:11:16 +00:00
var index uint64
var err error
if s.raftStore != nil {
index, err = s.raftStore.LastIndex()
} else if s.raftInmem != nil {
index, err = s.raftInmem.LastIndex()
} else {
panic("neither raftInmem or raftStore is initialized")
}
2015-06-04 11:11:35 +00:00
if err != nil {
s.logger.Printf("[ERR] nomad: failed to read last raft index: %v", err)
return
}
2015-06-03 10:58:00 +00:00
2015-06-04 11:11:35 +00:00
// Bootstrap can only be done if there are no committed logs,
// remove our expectations of bootstrapping
if index != 0 {
atomic.StoreInt32(&s.config.BootstrapExpect, 0)
2015-06-04 11:11:35 +00:00
return
}
2015-06-03 10:58:00 +00:00
2015-06-04 11:11:35 +00:00
// Scan for all the known servers
members := s.serf.Members()
addrs := make([]string, 0)
for _, member := range members {
valid, p := isNomadServer(member)
if !valid {
continue
}
if p.Region != s.config.Region {
continue
}
if p.Expect != 0 && p.Expect != int(atomic.LoadInt32(&s.config.BootstrapExpect)) {
2015-06-04 11:11:35 +00:00
s.logger.Printf("[ERR] nomad: peer %v has a conflicting expect value. All nodes should expect the same number.", member)
return
}
if p.Bootstrap {
s.logger.Printf("[ERR] nomad: peer %v has bootstrap mode. Expect disabled.", member)
return
}
addrs = append(addrs, p.Addr.String())
}
2015-06-03 10:58:00 +00:00
2015-06-04 11:11:35 +00:00
// Skip if we haven't met the minimum expect count
if len(addrs) < int(atomic.LoadInt32(&s.config.BootstrapExpect)) {
2015-06-04 11:11:35 +00:00
return
}
2015-06-03 10:58:00 +00:00
2015-06-04 11:11:35 +00:00
// Update the peer set
s.logger.Printf("[INFO] nomad: Attempting bootstrap with nodes: %v", addrs)
if err := s.raft.SetPeers(addrs).Error(); err != nil {
s.logger.Printf("[ERR] nomad: failed to bootstrap peers: %v", err)
}
2015-06-03 10:58:00 +00:00
// Bootstrapping complete, don't enter this again
atomic.StoreInt32(&s.config.BootstrapExpect, 0)
2015-06-03 10:58:00 +00:00
}
// nodeFailed is used to handle fail events on the serf cluster
func (s *Server) nodeFailed(me serf.MemberEvent) {
2015-06-04 11:02:39 +00:00
for _, m := range me.Members {
ok, parts := isNomadServer(m)
if !ok {
continue
}
s.logger.Printf("[INFO] nomad: removing server %s", parts)
2015-06-03 10:58:00 +00:00
2015-06-04 11:02:39 +00:00
// Remove the server if known
s.peerLock.Lock()
existing := s.peers[parts.Region]
n := len(existing)
for i := 0; i < n; i++ {
if existing[i].Name == parts.Name {
existing[i], existing[n-1] = existing[n-1], nil
existing = existing[:n-1]
n--
break
}
}
2015-06-03 10:58:00 +00:00
2015-06-04 11:02:39 +00:00
// Trim the list there are no known servers in a region
if n == 0 {
delete(s.peers, parts.Region)
} else {
s.peers[parts.Region] = existing
}
2015-06-07 18:37:59 +00:00
// Check if local peer
if parts.Region == s.config.Region {
delete(s.localPeers, parts.Addr.String())
}
2015-06-04 11:02:39 +00:00
s.peerLock.Unlock()
}
2015-06-03 10:58:00 +00:00
}
2015-06-04 10:42:56 +00:00
// localMemberEvent is used to reconcile Serf events with the
// consistent store if we are the current leader.
func (s *Server) localMemberEvent(me serf.MemberEvent) {
// Do nothing if we are not the leader
if !s.IsLeader() {
return
}
// Check if this is a reap event
isReap := me.EventType() == serf.EventMemberReap
// Queue the members for reconciliation
for _, m := range me.Members {
// Change the status if this is a reap event
if isReap {
m.Status = StatusReap
}
select {
case s.reconcileCh <- m:
default:
}
}
}