open-nomad/nomad/testing.go

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

189 lines
5.0 KiB
Go
Raw Normal View History

// Copyright (c) HashiCorp, Inc.
// SPDX-License-Identifier: MPL-2.0
2018-01-12 01:00:30 +00:00
package nomad
import (
"fmt"
"math/rand"
"net"
"sync/atomic"
"time"
"github.com/hashicorp/nomad/ci"
2018-01-12 01:00:30 +00:00
"github.com/hashicorp/nomad/command/agent/consul"
"github.com/hashicorp/nomad/helper/testlog"
"github.com/hashicorp/nomad/nomad/mock"
"github.com/hashicorp/nomad/nomad/structs"
2019-09-04 01:42:41 +00:00
"github.com/hashicorp/nomad/version"
testing "github.com/mitchellh/go-testing-interface"
"github.com/shoenig/test/must"
2018-01-12 01:00:30 +00:00
)
var (
nodeNumber int32 = 0
2018-01-12 01:00:30 +00:00
)
func TestACLServer(t testing.T, cb func(*Config)) (*Server, *structs.ACLToken, func()) {
server, cleanup := TestServer(t, func(c *Config) {
2018-01-12 01:00:30 +00:00
c.ACLEnabled = true
if cb != nil {
cb(c)
}
})
token := mock.ACLManagementToken()
err := server.State().BootstrapACLTokens(structs.MsgTypeTestSetup, 1, 0, token)
2018-01-12 01:00:30 +00:00
if err != nil {
t.Fatalf("failed to bootstrap ACL token: %v", err)
}
return server, token, cleanup
2018-01-12 01:00:30 +00:00
}
func TestServer(t testing.T, cb func(*Config)) (*Server, func()) {
s, c, err := TestServerErr(t, cb)
must.NoError(t, err, must.Sprint("failed to start test server"))
return s, c
}
// TestConfigForServer provides a fully functional Config to pass to NewServer()
// It can be changed beforehand to induce different behavior such as specific errors.
func TestConfigForServer(t testing.T) *Config {
t.Helper()
2018-01-12 01:00:30 +00:00
// Setup the default settings
config := DefaultConfig()
// Setup default enterprise-specific settings, including license
defaultEnterpriseTestConfig(config)
2019-09-04 01:42:41 +00:00
config.Build = version.Version + "+unittest"
2018-01-12 01:00:30 +00:00
config.DevMode = true
config.DataDir = t.TempDir()
config.EnableEventBroker = true
Simplify Bootstrap logic in tests This change updates tests to honor `BootstrapExpect` exclusively when forming test clusters and removes test only knobs, e.g. `config.DevDisableBootstrap`. Background: Test cluster creation is fragile. Test servers don't follow the BootstapExpected route like production clusters. Instead they start as single node clusters and then get rejoin and may risk causing brain split or other test flakiness. The test framework expose few knobs to control those (e.g. `config.DevDisableBootstrap` and `config.Bootstrap`) that control whether a server should bootstrap the cluster. These flags are confusing and it's unclear when to use: their usage in multi-node cluster isn't properly documented. Furthermore, they have some bad side-effects as they don't control Raft library: If `config.DevDisableBootstrap` is true, the test server may not immediately attempt to bootstrap a cluster, but after an election timeout (~50ms), Raft may force a leadership election and win it (with only one vote) and cause a split brain. The knobs are also confusing as Bootstrap is an overloaded term. In BootstrapExpect, we refer to bootstrapping the cluster only after N servers are connected. But in tests and the knobs above, it refers to whether the server is a single node cluster and shouldn't wait for any other server. Changes: This commit makes two changes: First, it relies on `BootstrapExpected` instead of `Bootstrap` and/or `DevMode` flags. This change is relatively trivial. Introduce a `Bootstrapped` flag to track if the cluster is bootstrapped. This allows us to keep `BootstrapExpected` immutable. Previously, the flag was a config value but it gets set to 0 after cluster bootstrap completes.
2020-03-02 15:29:24 +00:00
config.BootstrapExpect = 1
nodeNum := atomic.AddInt32(&nodeNumber, 1)
2018-01-12 01:00:30 +00:00
config.NodeName = fmt.Sprintf("nomad-%03d", nodeNum)
// configure logger
config.Logger, config.LogOutput = testlog.HCLoggerNode(t, nodeNum)
2018-01-12 01:00:30 +00:00
// Tighten the Serf timing
config.SerfConfig.MemberlistConfig.BindAddr = "127.0.0.1"
config.SerfConfig.MemberlistConfig.SuspicionMult = 2
config.SerfConfig.MemberlistConfig.RetransmitMult = 2
config.SerfConfig.MemberlistConfig.ProbeTimeout = 50 * time.Millisecond
config.SerfConfig.MemberlistConfig.ProbeInterval = 100 * time.Millisecond
config.SerfConfig.MemberlistConfig.GossipInterval = 100 * time.Millisecond
// Tighten the Raft timing
config.RaftConfig.LeaderLeaseTimeout = 50 * time.Millisecond
config.RaftConfig.HeartbeatTimeout = 50 * time.Millisecond
config.RaftConfig.ElectionTimeout = 50 * time.Millisecond
config.RaftTimeout = 500 * time.Millisecond
// Disable Vault
f := false
config.VaultConfig.Enabled = &f
2018-02-15 23:48:32 +00:00
// Tighten the autopilot timing
config.AutopilotConfig.ServerStabilizationTime = 100 * time.Millisecond
config.ServerHealthInterval = 50 * time.Millisecond
config.AutopilotInterval = 100 * time.Millisecond
// Disable consul autojoining: tests typically join servers directly
config.ConsulConfig.ServerAutoJoin = &f
// Enable fuzzy search API
config.SearchConfig = &structs.SearchConfig{
FuzzyEnabled: true,
LimitQuery: 20,
LimitResults: 100,
MinTermLength: 2,
}
// Get random ports for RPC and Serf
ports := ci.PortAllocator.Grab(2)
config.RPCAddr = &net.TCPAddr{
IP: []byte{127, 0, 0, 1},
Port: ports[0],
}
config.SerfConfig.MemberlistConfig.BindPort = ports[1]
// max job submission source size
config.JobMaxSourceSize = 1e6
return config
}
func TestServerErr(t testing.T, cb func(*Config)) (*Server, func(), error) {
config := TestConfigForServer(t)
2018-01-12 01:00:30 +00:00
// Invoke the callback if any
if cb != nil {
cb(config)
}
cCatalog := consul.NewMockCatalog(config.Logger)
cConfigs := consul.NewMockConfigsAPI(config.Logger)
cACLs := consul.NewMockACLsAPI(config.Logger)
var server *Server
var err error
2018-01-12 01:00:30 +00:00
for i := 10; i >= 0; i-- {
2018-01-12 01:00:30 +00:00
// Create server
server, err = NewServer(config, cCatalog, cConfigs, cACLs)
2018-01-12 01:00:30 +00:00
if err == nil {
return server, func() {
ch := make(chan error)
go func() {
defer close(ch)
// Shutdown server
err = server.Shutdown()
if err != nil {
2022-04-02 00:24:02 +00:00
ch <- fmt.Errorf("failed to shutdown server: %w", err)
}
}()
select {
case e := <-ch:
if e != nil {
t.Fatal(e.Error())
}
case <-time.After(1 * time.Minute):
t.Fatal("timed out while shutting down server")
}
}, nil
} else if i > 0 {
2018-01-12 01:00:30 +00:00
if server != nil {
_ = server.Shutdown()
2018-01-12 01:00:30 +00:00
}
wait := time.Duration(rand.Int31n(2000)) * time.Millisecond
time.Sleep(wait)
}
// if it failed for port reasons, try new ones
ports := ci.PortAllocator.Grab(2)
config.RPCAddr = &net.TCPAddr{
IP: []byte{127, 0, 0, 1},
Port: ports[0],
}
config.SerfConfig.MemberlistConfig.BindPort = ports[1]
2018-01-12 01:00:30 +00:00
}
return nil, nil, fmt.Errorf("error starting test server: %w", err)
2018-01-12 01:00:30 +00:00
}
func TestJoin(t testing.T, servers ...*Server) {
for i := 0; i < len(servers)-1; i++ {
addr := fmt.Sprintf("127.0.0.1:%d",
servers[i].config.SerfConfig.MemberlistConfig.BindPort)
for j := i + 1; j < len(servers); j++ {
num, err := servers[j].Join([]string{addr})
must.NoError(t, err)
must.Eq(t, 1, num)
2018-01-12 01:00:30 +00:00
}
}
}