blob: 78b74bfa583a5a764978d0cdb06c04443fb84d2e [file] [log] [blame]
// Copyright 2015 The etcd Authors
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package etcdserver
import (
"context"
"fmt"
"path/filepath"
"sort"
"strings"
"time"
"github.com/coreos/etcd/pkg/netutil"
"github.com/coreos/etcd/pkg/transport"
"github.com/coreos/etcd/pkg/types"
)
// ServerConfig holds the configuration of etcd as taken from the command line or discovery.
type ServerConfig struct {
Name string
DiscoveryURL string
DiscoveryProxy string
ClientURLs types.URLs
PeerURLs types.URLs
DataDir string
// DedicatedWALDir config will make the etcd to write the WAL to the WALDir
// rather than the dataDir/member/wal.
DedicatedWALDir string
SnapCount uint64
MaxSnapFiles uint
MaxWALFiles uint
InitialPeerURLsMap types.URLsMap
InitialClusterToken string
NewCluster bool
ForceNewCluster bool
PeerTLSInfo transport.TLSInfo
TickMs uint
ElectionTicks int
// InitialElectionTickAdvance is true, then local member fast-forwards
// election ticks to speed up "initial" leader election trigger. This
// benefits the case of larger election ticks. For instance, cross
// datacenter deployment may require longer election timeout of 10-second.
// If true, local node does not need wait up to 10-second. Instead,
// forwards its election ticks to 8-second, and have only 2-second left
// before leader election.
//
// Major assumptions are that:
// - cluster has no active leader thus advancing ticks enables faster
// leader election, or
// - cluster already has an established leader, and rejoining follower
// is likely to receive heartbeats from the leader after tick advance
// and before election timeout.
//
// However, when network from leader to rejoining follower is congested,
// and the follower does not receive leader heartbeat within left election
// ticks, disruptive election has to happen thus affecting cluster
// availabilities.
//
// Disabling this would slow down initial bootstrap process for cross
// datacenter deployments. Make your own tradeoffs by configuring
// --initial-election-tick-advance at the cost of slow initial bootstrap.
//
// If single-node, it advances ticks regardless.
//
// See https://github.com/coreos/etcd/issues/9333 for more detail.
InitialElectionTickAdvance bool
BootstrapTimeout time.Duration
AutoCompactionRetention time.Duration
AutoCompactionMode string
QuotaBackendBytes int64
MaxTxnOps uint
// MaxRequestBytes is the maximum request size to send over raft.
MaxRequestBytes uint
StrictReconfigCheck bool
// ClientCertAuthEnabled is true when cert has been signed by the client CA.
ClientCertAuthEnabled bool
AuthToken string
TokenTTL uint
// InitialCorruptCheck is true to check data corruption on boot
// before serving any peer/client traffic.
InitialCorruptCheck bool
CorruptCheckTime time.Duration
Debug bool
}
// VerifyBootstrap sanity-checks the initial config for bootstrap case
// and returns an error for things that should never happen.
func (c *ServerConfig) VerifyBootstrap() error {
if err := c.hasLocalMember(); err != nil {
return err
}
if err := c.advertiseMatchesCluster(); err != nil {
return err
}
if checkDuplicateURL(c.InitialPeerURLsMap) {
return fmt.Errorf("initial cluster %s has duplicate url", c.InitialPeerURLsMap)
}
if c.InitialPeerURLsMap.String() == "" && c.DiscoveryURL == "" {
return fmt.Errorf("initial cluster unset and no discovery URL found")
}
return nil
}
// VerifyJoinExisting sanity-checks the initial config for join existing cluster
// case and returns an error for things that should never happen.
func (c *ServerConfig) VerifyJoinExisting() error {
// The member has announced its peer urls to the cluster before starting; no need to
// set the configuration again.
if err := c.hasLocalMember(); err != nil {
return err
}
if checkDuplicateURL(c.InitialPeerURLsMap) {
return fmt.Errorf("initial cluster %s has duplicate url", c.InitialPeerURLsMap)
}
if c.DiscoveryURL != "" {
return fmt.Errorf("discovery URL should not be set when joining existing initial cluster")
}
return nil
}
// hasLocalMember checks that the cluster at least contains the local server.
func (c *ServerConfig) hasLocalMember() error {
if urls := c.InitialPeerURLsMap[c.Name]; urls == nil {
return fmt.Errorf("couldn't find local name %q in the initial cluster configuration", c.Name)
}
return nil
}
// advertiseMatchesCluster confirms peer URLs match those in the cluster peer list.
func (c *ServerConfig) advertiseMatchesCluster() error {
urls, apurls := c.InitialPeerURLsMap[c.Name], c.PeerURLs.StringSlice()
urls.Sort()
sort.Strings(apurls)
ctx, cancel := context.WithTimeout(context.TODO(), 30*time.Second)
defer cancel()
ok, err := netutil.URLStringsEqual(ctx, apurls, urls.StringSlice())
if ok {
return nil
}
initMap, apMap := make(map[string]struct{}), make(map[string]struct{})
for _, url := range c.PeerURLs {
apMap[url.String()] = struct{}{}
}
for _, url := range c.InitialPeerURLsMap[c.Name] {
initMap[url.String()] = struct{}{}
}
missing := []string{}
for url := range initMap {
if _, ok := apMap[url]; !ok {
missing = append(missing, url)
}
}
if len(missing) > 0 {
for i := range missing {
missing[i] = c.Name + "=" + missing[i]
}
mstr := strings.Join(missing, ",")
apStr := strings.Join(apurls, ",")
return fmt.Errorf("--initial-cluster has %s but missing from --initial-advertise-peer-urls=%s (%v)", mstr, apStr, err)
}
for url := range apMap {
if _, ok := initMap[url]; !ok {
missing = append(missing, url)
}
}
if len(missing) > 0 {
mstr := strings.Join(missing, ",")
umap := types.URLsMap(map[string]types.URLs{c.Name: c.PeerURLs})
return fmt.Errorf("--initial-advertise-peer-urls has %s but missing from --initial-cluster=%s", mstr, umap.String())
}
// resolved URLs from "--initial-advertise-peer-urls" and "--initial-cluster" did not match or failed
apStr := strings.Join(apurls, ",")
umap := types.URLsMap(map[string]types.URLs{c.Name: c.PeerURLs})
return fmt.Errorf("failed to resolve %s to match --initial-cluster=%s (%v)", apStr, umap.String(), err)
}
func (c *ServerConfig) MemberDir() string { return filepath.Join(c.DataDir, "member") }
func (c *ServerConfig) WALDir() string {
if c.DedicatedWALDir != "" {
return c.DedicatedWALDir
}
return filepath.Join(c.MemberDir(), "wal")
}
func (c *ServerConfig) SnapDir() string { return filepath.Join(c.MemberDir(), "snap") }
func (c *ServerConfig) ShouldDiscover() bool { return c.DiscoveryURL != "" }
// ReqTimeout returns timeout for request to finish.
func (c *ServerConfig) ReqTimeout() time.Duration {
// 5s for queue waiting, computation and disk IO delay
// + 2 * election timeout for possible leader election
return 5*time.Second + 2*time.Duration(c.ElectionTicks*int(c.TickMs))*time.Millisecond
}
func (c *ServerConfig) electionTimeout() time.Duration {
return time.Duration(c.ElectionTicks*int(c.TickMs)) * time.Millisecond
}
func (c *ServerConfig) peerDialTimeout() time.Duration {
// 1s for queue wait and election timeout
return time.Second + time.Duration(c.ElectionTicks*int(c.TickMs))*time.Millisecond
}
func (c *ServerConfig) PrintWithInitial() { c.print(true) }
func (c *ServerConfig) Print() { c.print(false) }
func (c *ServerConfig) print(initial bool) {
plog.Infof("name = %s", c.Name)
if c.ForceNewCluster {
plog.Infof("force new cluster")
}
plog.Infof("data dir = %s", c.DataDir)
plog.Infof("member dir = %s", c.MemberDir())
if c.DedicatedWALDir != "" {
plog.Infof("dedicated WAL dir = %s", c.DedicatedWALDir)
}
plog.Infof("heartbeat = %dms", c.TickMs)
plog.Infof("election = %dms", c.ElectionTicks*int(c.TickMs))
plog.Infof("snapshot count = %d", c.SnapCount)
if len(c.DiscoveryURL) != 0 {
plog.Infof("discovery URL= %s", c.DiscoveryURL)
if len(c.DiscoveryProxy) != 0 {
plog.Infof("discovery proxy = %s", c.DiscoveryProxy)
}
}
plog.Infof("advertise client URLs = %s", c.ClientURLs)
if initial {
plog.Infof("initial advertise peer URLs = %s", c.PeerURLs)
plog.Infof("initial cluster = %s", c.InitialPeerURLsMap)
}
}
func checkDuplicateURL(urlsmap types.URLsMap) bool {
um := make(map[string]bool)
for _, urls := range urlsmap {
for _, url := range urls {
u := url.String()
if um[u] {
return true
}
um[u] = true
}
}
return false
}
func (c *ServerConfig) bootstrapTimeout() time.Duration {
if c.BootstrapTimeout != 0 {
return c.BootstrapTimeout
}
return time.Second
}
func (c *ServerConfig) backendPath() string { return filepath.Join(c.SnapDir(), "db") }