69a088ca85
* new config parser for agent This patch implements a new config parser for the consul agent which makes the following changes to the previous implementation: * add HCL support * all configuration fragments in tests and for default config are expressed as HCL fragments * HCL fragments can be provided on the command line so that they can eventually replace the command line flags. * HCL/JSON fragments are parsed into a temporary Config structure which can be merged using reflection (all values are pointers). The existing merge logic of overwrite for values and append for slices has been preserved. * A single builder process generates a typed runtime configuration for the agent. The new implementation is more strict and fails in the builder process if no valid runtime configuration can be generated. Therefore, additional validations in other parts of the code should be removed. The builder also pre-computes all required network addresses so that no address/port magic should be required where the configuration is used and should therefore be removed. * Upgrade github.com/hashicorp/hcl to support int64 * improve error messages * fix directory permission test * Fix rtt test * Fix ForceLeave test * Skip performance test for now until we know what to do * Update github.com/hashicorp/memberlist to update log prefix * Make memberlist use the default logger * improve config error handling * do not fail on non-existing data-dir * experiment with non-uniform timeouts to get a handle on stalled leader elections * Run tests for packages separately to eliminate the spurious port conflicts * refactor private address detection and unify approach for ipv4 and ipv6. Fixes #2825 * do not allow unix sockets for DNS * improve bind and advertise addr error handling * go through builder using test coverage * minimal update to the docs * more coverage tests fixed * more tests * fix makefile * cleanup * fix port conflicts with external port server 'porter' * stop test server on error * do not run api test that change global ENV concurrently with the other tests * Run remaining api tests concurrently * no need for retry with the port number service * monkey patch race condition in go-sockaddr until we understand why that fails * monkey patch hcl decoder race condidtion until we understand why that fails * monkey patch spurious errors in strings.EqualFold from here * add test for hcl decoder race condition. Run with go test -parallel 128 * Increase timeout again * cleanup * don't log port allocations by default * use base command arg parsing to format help output properly * handle -dc deprecation case in Build * switch autopilot.max_trailing_logs to int * remove duplicate test case * remove unused methods * remove comments about flag/config value inconsistencies * switch got and want around since the error message was misleading. * Removes a stray debug log. * Removes a stray newline in imports. * Fixes TestACL_Version8. * Runs go fmt. * Adds a default case for unknown address types. * Reoders and reformats some imports. * Adds some comments and fixes typos. * Reorders imports. * add unix socket support for dns later * drop all deprecated flags and arguments * fix wrong field name * remove stray node-id file * drop unnecessary patch section in test * drop duplicate test * add test for LeaveOnTerm and SkipLeaveOnInt in client mode * drop "bla" and add clarifying comment for the test * split up tests to support enterprise/non-enterprise tests * drop raft multiplier and derive values during build phase * sanitize runtime config reflectively and add test * detect invalid config fields * fix tests with invalid config fields * use different values for wan sanitiziation test * drop recursor in favor of recursors * allow dns_config.udp_answer_limit to be zero * make sure tests run on machines with multiple ips * Fix failing tests in a few more places by providing a bind address in the test * Gets rid of skipped TestAgent_CheckPerformanceSettings and adds case for builder. * Add porter to server_test.go to make tests there less flaky * go fmt
1634 lines
39 KiB
Go
1634 lines
39 KiB
Go
package agent
|
|
|
|
import (
|
|
"reflect"
|
|
"testing"
|
|
"time"
|
|
|
|
"github.com/hashicorp/consul/agent/config"
|
|
"github.com/hashicorp/consul/agent/structs"
|
|
"github.com/hashicorp/consul/agent/token"
|
|
"github.com/hashicorp/consul/api"
|
|
"github.com/hashicorp/consul/testutil/retry"
|
|
"github.com/hashicorp/consul/types"
|
|
"github.com/pascaldekloe/goe/verify"
|
|
)
|
|
|
|
func TestAgentAntiEntropy_Services(t *testing.T) {
|
|
t.Parallel()
|
|
a := &TestAgent{Name: t.Name(), NoInitialSync: true}
|
|
a.Start()
|
|
defer a.Shutdown()
|
|
|
|
// Register info
|
|
args := &structs.RegisterRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
Address: "127.0.0.1",
|
|
}
|
|
|
|
// Exists both, same (noop)
|
|
var out struct{}
|
|
srv1 := &structs.NodeService{
|
|
ID: "mysql",
|
|
Service: "mysql",
|
|
Tags: []string{"master"},
|
|
Port: 5000,
|
|
}
|
|
a.state.AddService(srv1, "")
|
|
args.Service = srv1
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Exists both, different (update)
|
|
srv2 := &structs.NodeService{
|
|
ID: "redis",
|
|
Service: "redis",
|
|
Tags: []string{},
|
|
Port: 8000,
|
|
}
|
|
a.state.AddService(srv2, "")
|
|
|
|
srv2_mod := new(structs.NodeService)
|
|
*srv2_mod = *srv2
|
|
srv2_mod.Port = 9000
|
|
args.Service = srv2_mod
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Exists local (create)
|
|
srv3 := &structs.NodeService{
|
|
ID: "web",
|
|
Service: "web",
|
|
Tags: []string{},
|
|
Port: 80,
|
|
}
|
|
a.state.AddService(srv3, "")
|
|
|
|
// Exists remote (delete)
|
|
srv4 := &structs.NodeService{
|
|
ID: "lb",
|
|
Service: "lb",
|
|
Tags: []string{},
|
|
Port: 443,
|
|
}
|
|
args.Service = srv4
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Exists both, different address (update)
|
|
srv5 := &structs.NodeService{
|
|
ID: "api",
|
|
Service: "api",
|
|
Tags: []string{},
|
|
Address: "127.0.0.10",
|
|
Port: 8000,
|
|
}
|
|
a.state.AddService(srv5, "")
|
|
|
|
srv5_mod := new(structs.NodeService)
|
|
*srv5_mod = *srv5
|
|
srv5_mod.Address = "127.0.0.1"
|
|
args.Service = srv5_mod
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Exists local, in sync, remote missing (create)
|
|
srv6 := &structs.NodeService{
|
|
ID: "cache",
|
|
Service: "cache",
|
|
Tags: []string{},
|
|
Port: 11211,
|
|
}
|
|
a.state.AddService(srv6, "")
|
|
|
|
// todo(fs): data race
|
|
a.state.Lock()
|
|
a.state.serviceStatus["cache"] = syncStatus{inSync: true}
|
|
a.state.Unlock()
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
|
|
var services structs.IndexedNodeServices
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
}
|
|
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Make sure we sent along our node info when we synced.
|
|
id := services.NodeServices.Node.ID
|
|
addrs := services.NodeServices.Node.TaggedAddresses
|
|
meta := services.NodeServices.Node.Meta
|
|
delete(meta, structs.MetaSegmentKey) // Added later, not in config.
|
|
verify.Values(r, "node id", id, a.config.NodeID)
|
|
verify.Values(r, "tagged addrs", addrs, a.config.TaggedAddresses)
|
|
verify.Values(r, "node meta", meta, a.config.NodeMeta)
|
|
|
|
// We should have 6 services (consul included)
|
|
if len(services.NodeServices.Services) != 6 {
|
|
r.Fatalf("bad: %v", services.NodeServices.Services)
|
|
}
|
|
|
|
// All the services should match
|
|
for id, serv := range services.NodeServices.Services {
|
|
serv.CreateIndex, serv.ModifyIndex = 0, 0
|
|
switch id {
|
|
case "mysql":
|
|
if !reflect.DeepEqual(serv, srv1) {
|
|
r.Fatalf("bad: %v %v", serv, srv1)
|
|
}
|
|
case "redis":
|
|
if !reflect.DeepEqual(serv, srv2) {
|
|
r.Fatalf("bad: %#v %#v", serv, srv2)
|
|
}
|
|
case "web":
|
|
if !reflect.DeepEqual(serv, srv3) {
|
|
r.Fatalf("bad: %v %v", serv, srv3)
|
|
}
|
|
case "api":
|
|
if !reflect.DeepEqual(serv, srv5) {
|
|
r.Fatalf("bad: %v %v", serv, srv5)
|
|
}
|
|
case "cache":
|
|
if !reflect.DeepEqual(serv, srv6) {
|
|
r.Fatalf("bad: %v %v", serv, srv6)
|
|
}
|
|
case structs.ConsulServiceID:
|
|
// ignore
|
|
default:
|
|
r.Fatalf("unexpected service: %v", id)
|
|
}
|
|
}
|
|
|
|
// todo(fs): data race
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state
|
|
if len(a.state.services) != 5 {
|
|
r.Fatalf("bad: %v", a.state.services)
|
|
}
|
|
if len(a.state.serviceStatus) != 5 {
|
|
r.Fatalf("bad: %v", a.state.serviceStatus)
|
|
}
|
|
for name, status := range a.state.serviceStatus {
|
|
if !status.inSync {
|
|
r.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
})
|
|
|
|
// Remove one of the services
|
|
a.state.RemoveService("api")
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// We should have 5 services (consul included)
|
|
if len(services.NodeServices.Services) != 5 {
|
|
r.Fatalf("bad: %v", services.NodeServices.Services)
|
|
}
|
|
|
|
// All the services should match
|
|
for id, serv := range services.NodeServices.Services {
|
|
serv.CreateIndex, serv.ModifyIndex = 0, 0
|
|
switch id {
|
|
case "mysql":
|
|
if !reflect.DeepEqual(serv, srv1) {
|
|
r.Fatalf("bad: %v %v", serv, srv1)
|
|
}
|
|
case "redis":
|
|
if !reflect.DeepEqual(serv, srv2) {
|
|
r.Fatalf("bad: %#v %#v", serv, srv2)
|
|
}
|
|
case "web":
|
|
if !reflect.DeepEqual(serv, srv3) {
|
|
r.Fatalf("bad: %v %v", serv, srv3)
|
|
}
|
|
case "cache":
|
|
if !reflect.DeepEqual(serv, srv6) {
|
|
r.Fatalf("bad: %v %v", serv, srv6)
|
|
}
|
|
case structs.ConsulServiceID:
|
|
// ignore
|
|
default:
|
|
r.Fatalf("unexpected service: %v", id)
|
|
}
|
|
}
|
|
|
|
// todo(fs): data race
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state
|
|
if len(a.state.services) != 4 {
|
|
r.Fatalf("bad: %v", a.state.services)
|
|
}
|
|
if len(a.state.serviceStatus) != 4 {
|
|
r.Fatalf("bad: %v", a.state.serviceStatus)
|
|
}
|
|
for name, status := range a.state.serviceStatus {
|
|
if !status.inSync {
|
|
r.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
})
|
|
}
|
|
|
|
func TestAgentAntiEntropy_EnableTagOverride(t *testing.T) {
|
|
t.Parallel()
|
|
a := &TestAgent{Name: t.Name(), NoInitialSync: true}
|
|
a.Start()
|
|
defer a.Shutdown()
|
|
|
|
args := &structs.RegisterRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
Address: "127.0.0.1",
|
|
}
|
|
var out struct{}
|
|
|
|
// EnableTagOverride = true
|
|
srv1 := &structs.NodeService{
|
|
ID: "svc_id1",
|
|
Service: "svc1",
|
|
Tags: []string{"tag1"},
|
|
Port: 6100,
|
|
EnableTagOverride: true,
|
|
}
|
|
a.state.AddService(srv1, "")
|
|
srv1_mod := new(structs.NodeService)
|
|
*srv1_mod = *srv1
|
|
srv1_mod.Port = 7100
|
|
srv1_mod.Tags = []string{"tag1_mod"}
|
|
args.Service = srv1_mod
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// EnableTagOverride = false
|
|
srv2 := &structs.NodeService{
|
|
ID: "svc_id2",
|
|
Service: "svc2",
|
|
Tags: []string{"tag2"},
|
|
Port: 6200,
|
|
EnableTagOverride: false,
|
|
}
|
|
a.state.AddService(srv2, "")
|
|
srv2_mod := new(structs.NodeService)
|
|
*srv2_mod = *srv2
|
|
srv2_mod.Port = 7200
|
|
srv2_mod.Tags = []string{"tag2_mod"}
|
|
args.Service = srv2_mod
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
}
|
|
var services structs.IndexedNodeServices
|
|
|
|
retry.Run(t, func(r *retry.R) {
|
|
// runtime.Gosched()
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// All the services should match
|
|
for id, serv := range services.NodeServices.Services {
|
|
serv.CreateIndex, serv.ModifyIndex = 0, 0
|
|
switch id {
|
|
case "svc_id1":
|
|
if serv.ID != "svc_id1" ||
|
|
serv.Service != "svc1" ||
|
|
serv.Port != 6100 ||
|
|
!reflect.DeepEqual(serv.Tags, []string{"tag1_mod"}) {
|
|
r.Fatalf("bad: %v %v", serv, srv1)
|
|
}
|
|
case "svc_id2":
|
|
if serv.ID != "svc_id2" ||
|
|
serv.Service != "svc2" ||
|
|
serv.Port != 6200 ||
|
|
!reflect.DeepEqual(serv.Tags, []string{"tag2"}) {
|
|
r.Fatalf("bad: %v %v", serv, srv2)
|
|
}
|
|
case structs.ConsulServiceID:
|
|
// ignore
|
|
default:
|
|
r.Fatalf("unexpected service: %v", id)
|
|
}
|
|
}
|
|
|
|
// todo(fs): data race
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
for name, status := range a.state.serviceStatus {
|
|
if !status.inSync {
|
|
r.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
})
|
|
}
|
|
|
|
func TestAgentAntiEntropy_Services_WithChecks(t *testing.T) {
|
|
t.Parallel()
|
|
a := NewTestAgent(t.Name(), "")
|
|
defer a.Shutdown()
|
|
|
|
{
|
|
// Single check
|
|
srv := &structs.NodeService{
|
|
ID: "mysql",
|
|
Service: "mysql",
|
|
Tags: []string{"master"},
|
|
Port: 5000,
|
|
}
|
|
a.state.AddService(srv, "")
|
|
|
|
chk := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "mysql",
|
|
Name: "mysql",
|
|
ServiceID: "mysql",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk, "")
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Sync the service once
|
|
if err := a.state.syncService("mysql"); err != nil {
|
|
t.Fatalf("err: %s", err)
|
|
}
|
|
}()
|
|
|
|
// We should have 2 services (consul included)
|
|
svcReq := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
}
|
|
var services structs.IndexedNodeServices
|
|
if err := a.RPC("Catalog.NodeServices", &svcReq, &services); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
if len(services.NodeServices.Services) != 2 {
|
|
t.Fatalf("bad: %v", services.NodeServices.Services)
|
|
}
|
|
|
|
// We should have one health check
|
|
chkReq := structs.ServiceSpecificRequest{
|
|
Datacenter: "dc1",
|
|
ServiceName: "mysql",
|
|
}
|
|
var checks structs.IndexedHealthChecks
|
|
if err := a.RPC("Health.ServiceChecks", &chkReq, &checks); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
if len(checks.HealthChecks) != 1 {
|
|
t.Fatalf("bad: %v", checks)
|
|
}
|
|
}
|
|
|
|
{
|
|
// Multiple checks
|
|
srv := &structs.NodeService{
|
|
ID: "redis",
|
|
Service: "redis",
|
|
Tags: []string{"master"},
|
|
Port: 5000,
|
|
}
|
|
a.state.AddService(srv, "")
|
|
|
|
chk1 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "redis:1",
|
|
Name: "redis:1",
|
|
ServiceID: "redis",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk1, "")
|
|
|
|
chk2 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "redis:2",
|
|
Name: "redis:2",
|
|
ServiceID: "redis",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk2, "")
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Sync the service once
|
|
if err := a.state.syncService("redis"); err != nil {
|
|
t.Fatalf("err: %s", err)
|
|
}
|
|
}()
|
|
|
|
// We should have 3 services (consul included)
|
|
svcReq := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
}
|
|
var services structs.IndexedNodeServices
|
|
if err := a.RPC("Catalog.NodeServices", &svcReq, &services); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
if len(services.NodeServices.Services) != 3 {
|
|
t.Fatalf("bad: %v", services.NodeServices.Services)
|
|
}
|
|
|
|
// We should have two health checks
|
|
chkReq := structs.ServiceSpecificRequest{
|
|
Datacenter: "dc1",
|
|
ServiceName: "redis",
|
|
}
|
|
var checks structs.IndexedHealthChecks
|
|
if err := a.RPC("Health.ServiceChecks", &chkReq, &checks); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
if len(checks.HealthChecks) != 2 {
|
|
t.Fatalf("bad: %v", checks)
|
|
}
|
|
}
|
|
}
|
|
|
|
var testRegisterRules = `
|
|
node "" {
|
|
policy = "write"
|
|
}
|
|
|
|
service "api" {
|
|
policy = "write"
|
|
}
|
|
|
|
service "consul" {
|
|
policy = "write"
|
|
}
|
|
`
|
|
|
|
func TestAgentAntiEntropy_Services_ACLDeny(t *testing.T) {
|
|
t.Parallel()
|
|
a := &TestAgent{Name: t.Name(), HCL: `
|
|
acl_datacenter = "dc1"
|
|
acl_master_token = "root"
|
|
acl_default_policy = "deny"
|
|
acl_enforce_version_8 = true
|
|
`, NoInitialSync: true}
|
|
a.Start()
|
|
defer a.Shutdown()
|
|
|
|
// Create the ACL
|
|
arg := structs.ACLRequest{
|
|
Datacenter: "dc1",
|
|
Op: structs.ACLSet,
|
|
ACL: structs.ACL{
|
|
Name: "User token",
|
|
Type: structs.ACLTypeClient,
|
|
Rules: testRegisterRules,
|
|
},
|
|
WriteRequest: structs.WriteRequest{
|
|
Token: "root",
|
|
},
|
|
}
|
|
var token string
|
|
if err := a.RPC("ACL.Apply", &arg, &token); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Create service (disallowed)
|
|
srv1 := &structs.NodeService{
|
|
ID: "mysql",
|
|
Service: "mysql",
|
|
Tags: []string{"master"},
|
|
Port: 5000,
|
|
}
|
|
a.state.AddService(srv1, token)
|
|
|
|
// Create service (allowed)
|
|
srv2 := &structs.NodeService{
|
|
ID: "api",
|
|
Service: "api",
|
|
Tags: []string{"foo"},
|
|
Port: 5001,
|
|
}
|
|
a.state.AddService(srv2, token)
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
time.Sleep(200 * time.Millisecond)
|
|
|
|
// Verify that we are in sync
|
|
{
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
QueryOptions: structs.QueryOptions{
|
|
Token: "root",
|
|
},
|
|
}
|
|
var services structs.IndexedNodeServices
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// We should have 2 services (consul included)
|
|
if len(services.NodeServices.Services) != 2 {
|
|
t.Fatalf("bad: %v", services.NodeServices.Services)
|
|
}
|
|
|
|
// All the services should match
|
|
for id, serv := range services.NodeServices.Services {
|
|
serv.CreateIndex, serv.ModifyIndex = 0, 0
|
|
switch id {
|
|
case "mysql":
|
|
t.Fatalf("should not be permitted")
|
|
case "api":
|
|
if !reflect.DeepEqual(serv, srv2) {
|
|
t.Fatalf("bad: %#v %#v", serv, srv2)
|
|
}
|
|
case structs.ConsulServiceID:
|
|
// ignore
|
|
default:
|
|
t.Fatalf("unexpected service: %v", id)
|
|
}
|
|
}
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state
|
|
if len(a.state.services) != 2 {
|
|
t.Fatalf("bad: %v", a.state.services)
|
|
}
|
|
if len(a.state.serviceStatus) != 2 {
|
|
t.Fatalf("bad: %v", a.state.serviceStatus)
|
|
}
|
|
for name, status := range a.state.serviceStatus {
|
|
if !status.inSync {
|
|
t.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
}()
|
|
}
|
|
|
|
// Now remove the service and re-sync
|
|
a.state.RemoveService("api")
|
|
a.StartSync()
|
|
time.Sleep(200 * time.Millisecond)
|
|
|
|
// Verify that we are in sync
|
|
{
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
QueryOptions: structs.QueryOptions{
|
|
Token: "root",
|
|
},
|
|
}
|
|
var services structs.IndexedNodeServices
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// We should have 1 service (just consul)
|
|
if len(services.NodeServices.Services) != 1 {
|
|
t.Fatalf("bad: %v", services.NodeServices.Services)
|
|
}
|
|
|
|
// All the services should match
|
|
for id, serv := range services.NodeServices.Services {
|
|
serv.CreateIndex, serv.ModifyIndex = 0, 0
|
|
switch id {
|
|
case "mysql":
|
|
t.Fatalf("should not be permitted")
|
|
case "api":
|
|
t.Fatalf("should be deleted")
|
|
case structs.ConsulServiceID:
|
|
// ignore
|
|
default:
|
|
t.Fatalf("unexpected service: %v", id)
|
|
}
|
|
}
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state
|
|
if len(a.state.services) != 1 {
|
|
t.Fatalf("bad: %v", a.state.services)
|
|
}
|
|
if len(a.state.serviceStatus) != 1 {
|
|
t.Fatalf("bad: %v", a.state.serviceStatus)
|
|
}
|
|
for name, status := range a.state.serviceStatus {
|
|
if !status.inSync {
|
|
t.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
}()
|
|
}
|
|
|
|
// Make sure the token got cleaned up.
|
|
if token := a.state.ServiceToken("api"); token != "" {
|
|
t.Fatalf("bad: %s", token)
|
|
}
|
|
}
|
|
|
|
func TestAgentAntiEntropy_Checks(t *testing.T) {
|
|
t.Parallel()
|
|
a := &TestAgent{Name: t.Name(), NoInitialSync: true}
|
|
a.Start()
|
|
defer a.Shutdown()
|
|
|
|
// Register info
|
|
args := &structs.RegisterRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
Address: "127.0.0.1",
|
|
}
|
|
|
|
// Exists both, same (noop)
|
|
var out struct{}
|
|
chk1 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "mysql",
|
|
Name: "mysql",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk1, "")
|
|
args.Check = chk1
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Exists both, different (update)
|
|
chk2 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "redis",
|
|
Name: "redis",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk2, "")
|
|
|
|
chk2_mod := new(structs.HealthCheck)
|
|
*chk2_mod = *chk2
|
|
chk2_mod.Status = api.HealthCritical
|
|
args.Check = chk2_mod
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Exists local (create)
|
|
chk3 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "web",
|
|
Name: "web",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk3, "")
|
|
|
|
// Exists remote (delete)
|
|
chk4 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "lb",
|
|
Name: "lb",
|
|
Status: api.HealthPassing,
|
|
}
|
|
args.Check = chk4
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Exists local, in sync, remote missing (create)
|
|
chk5 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "cache",
|
|
Name: "cache",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk5, "")
|
|
|
|
// todo(fs): data race
|
|
a.state.Lock()
|
|
a.state.checkStatus["cache"] = syncStatus{inSync: true}
|
|
a.state.Unlock()
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
}
|
|
var checks structs.IndexedHealthChecks
|
|
|
|
// Verify that we are in sync
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// We should have 5 checks (serf included)
|
|
if len(checks.HealthChecks) != 5 {
|
|
r.Fatalf("bad: %v", checks)
|
|
}
|
|
|
|
// All the checks should match
|
|
for _, chk := range checks.HealthChecks {
|
|
chk.CreateIndex, chk.ModifyIndex = 0, 0
|
|
switch chk.CheckID {
|
|
case "mysql":
|
|
if !reflect.DeepEqual(chk, chk1) {
|
|
r.Fatalf("bad: %v %v", chk, chk1)
|
|
}
|
|
case "redis":
|
|
if !reflect.DeepEqual(chk, chk2) {
|
|
r.Fatalf("bad: %v %v", chk, chk2)
|
|
}
|
|
case "web":
|
|
if !reflect.DeepEqual(chk, chk3) {
|
|
r.Fatalf("bad: %v %v", chk, chk3)
|
|
}
|
|
case "cache":
|
|
if !reflect.DeepEqual(chk, chk5) {
|
|
r.Fatalf("bad: %v %v", chk, chk5)
|
|
}
|
|
case "serfHealth":
|
|
// ignore
|
|
default:
|
|
r.Fatalf("unexpected check: %v", chk)
|
|
}
|
|
}
|
|
})
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state
|
|
if len(a.state.checks) != 4 {
|
|
t.Fatalf("bad: %v", a.state.checks)
|
|
}
|
|
if len(a.state.checkStatus) != 4 {
|
|
t.Fatalf("bad: %v", a.state.checkStatus)
|
|
}
|
|
for name, status := range a.state.checkStatus {
|
|
if !status.inSync {
|
|
t.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
}()
|
|
|
|
// Make sure we sent along our node info addresses when we synced.
|
|
{
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
}
|
|
var services structs.IndexedNodeServices
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
id := services.NodeServices.Node.ID
|
|
addrs := services.NodeServices.Node.TaggedAddresses
|
|
meta := services.NodeServices.Node.Meta
|
|
delete(meta, structs.MetaSegmentKey) // Added later, not in config.
|
|
verify.Values(t, "node id", id, a.config.NodeID)
|
|
verify.Values(t, "tagged addrs", addrs, a.config.TaggedAddresses)
|
|
verify.Values(t, "node meta", meta, a.config.NodeMeta)
|
|
}
|
|
|
|
// Remove one of the checks
|
|
a.state.RemoveCheck("redis")
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
|
|
// Verify that we are in sync
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// We should have 5 checks (serf included)
|
|
if len(checks.HealthChecks) != 4 {
|
|
r.Fatalf("bad: %v", checks)
|
|
}
|
|
|
|
// All the checks should match
|
|
for _, chk := range checks.HealthChecks {
|
|
chk.CreateIndex, chk.ModifyIndex = 0, 0
|
|
switch chk.CheckID {
|
|
case "mysql":
|
|
if !reflect.DeepEqual(chk, chk1) {
|
|
r.Fatalf("bad: %v %v", chk, chk1)
|
|
}
|
|
case "web":
|
|
if !reflect.DeepEqual(chk, chk3) {
|
|
r.Fatalf("bad: %v %v", chk, chk3)
|
|
}
|
|
case "cache":
|
|
if !reflect.DeepEqual(chk, chk5) {
|
|
r.Fatalf("bad: %v %v", chk, chk5)
|
|
}
|
|
case "serfHealth":
|
|
// ignore
|
|
default:
|
|
r.Fatalf("unexpected check: %v", chk)
|
|
}
|
|
}
|
|
})
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state
|
|
if len(a.state.checks) != 3 {
|
|
t.Fatalf("bad: %v", a.state.checks)
|
|
}
|
|
if len(a.state.checkStatus) != 3 {
|
|
t.Fatalf("bad: %v", a.state.checkStatus)
|
|
}
|
|
for name, status := range a.state.checkStatus {
|
|
if !status.inSync {
|
|
t.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
}()
|
|
}
|
|
|
|
func TestAgentAntiEntropy_Checks_ACLDeny(t *testing.T) {
|
|
t.Parallel()
|
|
a := &TestAgent{Name: t.Name(), HCL: `
|
|
acl_datacenter = "dc1"
|
|
acl_master_token = "root"
|
|
acl_default_policy = "deny"
|
|
acl_enforce_version_8 = true
|
|
`, NoInitialSync: true}
|
|
a.Start()
|
|
defer a.Shutdown()
|
|
|
|
// Create the ACL
|
|
arg := structs.ACLRequest{
|
|
Datacenter: "dc1",
|
|
Op: structs.ACLSet,
|
|
ACL: structs.ACL{
|
|
Name: "User token",
|
|
Type: structs.ACLTypeClient,
|
|
Rules: testRegisterRules,
|
|
},
|
|
WriteRequest: structs.WriteRequest{
|
|
Token: "root",
|
|
},
|
|
}
|
|
var token string
|
|
if err := a.RPC("ACL.Apply", &arg, &token); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Create services using the root token
|
|
srv1 := &structs.NodeService{
|
|
ID: "mysql",
|
|
Service: "mysql",
|
|
Tags: []string{"master"},
|
|
Port: 5000,
|
|
}
|
|
a.state.AddService(srv1, "root")
|
|
srv2 := &structs.NodeService{
|
|
ID: "api",
|
|
Service: "api",
|
|
Tags: []string{"foo"},
|
|
Port: 5001,
|
|
}
|
|
a.state.AddService(srv2, "root")
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
time.Sleep(200 * time.Millisecond)
|
|
|
|
// Verify that we are in sync
|
|
{
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
QueryOptions: structs.QueryOptions{
|
|
Token: "root",
|
|
},
|
|
}
|
|
var services structs.IndexedNodeServices
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// We should have 3 services (consul included)
|
|
if len(services.NodeServices.Services) != 3 {
|
|
t.Fatalf("bad: %v", services.NodeServices.Services)
|
|
}
|
|
|
|
// All the services should match
|
|
for id, serv := range services.NodeServices.Services {
|
|
serv.CreateIndex, serv.ModifyIndex = 0, 0
|
|
switch id {
|
|
case "mysql":
|
|
if !reflect.DeepEqual(serv, srv1) {
|
|
t.Fatalf("bad: %#v %#v", serv, srv1)
|
|
}
|
|
case "api":
|
|
if !reflect.DeepEqual(serv, srv2) {
|
|
t.Fatalf("bad: %#v %#v", serv, srv2)
|
|
}
|
|
case structs.ConsulServiceID:
|
|
// ignore
|
|
default:
|
|
t.Fatalf("unexpected service: %v", id)
|
|
}
|
|
}
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state
|
|
if len(a.state.services) != 2 {
|
|
t.Fatalf("bad: %v", a.state.services)
|
|
}
|
|
if len(a.state.serviceStatus) != 2 {
|
|
t.Fatalf("bad: %v", a.state.serviceStatus)
|
|
}
|
|
for name, status := range a.state.serviceStatus {
|
|
if !status.inSync {
|
|
t.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
}()
|
|
}
|
|
|
|
// This check won't be allowed.
|
|
chk1 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
ServiceID: "mysql",
|
|
ServiceName: "mysql",
|
|
ServiceTags: []string{"master"},
|
|
CheckID: "mysql-check",
|
|
Name: "mysql",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk1, token)
|
|
|
|
// This one will be allowed.
|
|
chk2 := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
ServiceID: "api",
|
|
ServiceName: "api",
|
|
ServiceTags: []string{"foo"},
|
|
CheckID: "api-check",
|
|
Name: "api",
|
|
Status: api.HealthPassing,
|
|
}
|
|
a.state.AddCheck(chk2, token)
|
|
|
|
// Trigger anti-entropy run and wait.
|
|
a.StartSync()
|
|
time.Sleep(200 * time.Millisecond)
|
|
|
|
// Verify that we are in sync
|
|
retry.Run(t, func(r *retry.R) {
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
QueryOptions: structs.QueryOptions{
|
|
Token: "root",
|
|
},
|
|
}
|
|
var checks structs.IndexedHealthChecks
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// We should have 2 checks (serf included)
|
|
if len(checks.HealthChecks) != 2 {
|
|
r.Fatalf("bad: %v", checks)
|
|
}
|
|
|
|
// All the checks should match
|
|
for _, chk := range checks.HealthChecks {
|
|
chk.CreateIndex, chk.ModifyIndex = 0, 0
|
|
switch chk.CheckID {
|
|
case "mysql-check":
|
|
t.Fatalf("should not be permitted")
|
|
case "api-check":
|
|
if !reflect.DeepEqual(chk, chk2) {
|
|
r.Fatalf("bad: %v %v", chk, chk2)
|
|
}
|
|
case "serfHealth":
|
|
// ignore
|
|
default:
|
|
r.Fatalf("unexpected check: %v", chk)
|
|
}
|
|
}
|
|
})
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state.
|
|
if len(a.state.checks) != 2 {
|
|
t.Fatalf("bad: %v", a.state.checks)
|
|
}
|
|
if len(a.state.checkStatus) != 2 {
|
|
t.Fatalf("bad: %v", a.state.checkStatus)
|
|
}
|
|
for name, status := range a.state.checkStatus {
|
|
if !status.inSync {
|
|
t.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
}()
|
|
|
|
// Now delete the check and wait for sync.
|
|
a.state.RemoveCheck("api-check")
|
|
a.StartSync()
|
|
time.Sleep(200 * time.Millisecond)
|
|
// Verify that we are in sync
|
|
retry.Run(t, func(r *retry.R) {
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
QueryOptions: structs.QueryOptions{
|
|
Token: "root",
|
|
},
|
|
}
|
|
var checks structs.IndexedHealthChecks
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// We should have 1 check (just serf)
|
|
if len(checks.HealthChecks) != 1 {
|
|
r.Fatalf("bad: %v", checks)
|
|
}
|
|
|
|
// All the checks should match
|
|
for _, chk := range checks.HealthChecks {
|
|
chk.CreateIndex, chk.ModifyIndex = 0, 0
|
|
switch chk.CheckID {
|
|
case "mysql-check":
|
|
r.Fatalf("should not be permitted")
|
|
case "api-check":
|
|
r.Fatalf("should be deleted")
|
|
case "serfHealth":
|
|
// ignore
|
|
default:
|
|
r.Fatalf("unexpected check: %v", chk)
|
|
}
|
|
}
|
|
})
|
|
|
|
// todo(fs): data race
|
|
func() {
|
|
a.state.RLock()
|
|
defer a.state.RUnlock()
|
|
|
|
// Check the local state.
|
|
if len(a.state.checks) != 1 {
|
|
t.Fatalf("bad: %v", a.state.checks)
|
|
}
|
|
if len(a.state.checkStatus) != 1 {
|
|
t.Fatalf("bad: %v", a.state.checkStatus)
|
|
}
|
|
for name, status := range a.state.checkStatus {
|
|
if !status.inSync {
|
|
t.Fatalf("should be in sync: %v %v", name, status)
|
|
}
|
|
}
|
|
}()
|
|
|
|
// Make sure the token got cleaned up.
|
|
if token := a.state.CheckToken("api-check"); token != "" {
|
|
t.Fatalf("bad: %s", token)
|
|
}
|
|
}
|
|
|
|
func TestAgentAntiEntropy_Check_DeferSync(t *testing.T) {
|
|
t.Parallel()
|
|
a := &TestAgent{Name: t.Name(), HCL: `
|
|
check_update_interval = "500ms"
|
|
`, NoInitialSync: true}
|
|
a.Start()
|
|
defer a.Shutdown()
|
|
|
|
// Create a check
|
|
check := &structs.HealthCheck{
|
|
Node: a.Config.NodeName,
|
|
CheckID: "web",
|
|
Name: "web",
|
|
Status: api.HealthPassing,
|
|
Output: "",
|
|
}
|
|
a.state.AddCheck(check, "")
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
|
|
// Verify that we are in sync
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
}
|
|
var checks structs.IndexedHealthChecks
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
if got, want := len(checks.HealthChecks), 2; got != want {
|
|
r.Fatalf("got %d health checks want %d", got, want)
|
|
}
|
|
})
|
|
|
|
// Update the check output! Should be deferred
|
|
a.state.UpdateCheck("web", api.HealthPassing, "output")
|
|
|
|
// Should not update for 500 milliseconds
|
|
time.Sleep(250 * time.Millisecond)
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Verify not updated
|
|
for _, chk := range checks.HealthChecks {
|
|
switch chk.CheckID {
|
|
case "web":
|
|
if chk.Output != "" {
|
|
t.Fatalf("early update: %v", chk)
|
|
}
|
|
}
|
|
}
|
|
// Wait for a deferred update
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
r.Fatal(err)
|
|
}
|
|
|
|
// Verify updated
|
|
for _, chk := range checks.HealthChecks {
|
|
switch chk.CheckID {
|
|
case "web":
|
|
if chk.Output != "output" {
|
|
r.Fatalf("no update: %v", chk)
|
|
}
|
|
}
|
|
}
|
|
})
|
|
|
|
// Change the output in the catalog to force it out of sync.
|
|
eCopy := check.Clone()
|
|
eCopy.Output = "changed"
|
|
reg := structs.RegisterRequest{
|
|
Datacenter: a.Config.Datacenter,
|
|
Node: a.Config.NodeName,
|
|
Address: a.Config.AdvertiseAddrLAN.IP.String(),
|
|
TaggedAddresses: a.Config.TaggedAddresses,
|
|
Check: eCopy,
|
|
WriteRequest: structs.WriteRequest{},
|
|
}
|
|
var out struct{}
|
|
if err := a.RPC("Catalog.Register", ®, &out); err != nil {
|
|
t.Fatalf("err: %s", err)
|
|
}
|
|
|
|
// Verify that the output is out of sync.
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
for _, chk := range checks.HealthChecks {
|
|
switch chk.CheckID {
|
|
case "web":
|
|
if chk.Output != "changed" {
|
|
t.Fatalf("unexpected update: %v", chk)
|
|
}
|
|
}
|
|
}
|
|
|
|
// Trigger anti-entropy run and wait.
|
|
a.StartSync()
|
|
time.Sleep(200 * time.Millisecond)
|
|
|
|
// Verify that the output was synced back to the agent's value.
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
for _, chk := range checks.HealthChecks {
|
|
switch chk.CheckID {
|
|
case "web":
|
|
if chk.Output != "output" {
|
|
t.Fatalf("missed update: %v", chk)
|
|
}
|
|
}
|
|
}
|
|
|
|
// Reset the catalog again.
|
|
if err := a.RPC("Catalog.Register", ®, &out); err != nil {
|
|
t.Fatalf("err: %s", err)
|
|
}
|
|
|
|
// Verify that the output is out of sync.
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
for _, chk := range checks.HealthChecks {
|
|
switch chk.CheckID {
|
|
case "web":
|
|
if chk.Output != "changed" {
|
|
t.Fatalf("unexpected update: %v", chk)
|
|
}
|
|
}
|
|
}
|
|
|
|
// Now make an update that should be deferred.
|
|
a.state.UpdateCheck("web", api.HealthPassing, "deferred")
|
|
|
|
// Trigger anti-entropy run and wait.
|
|
a.StartSync()
|
|
time.Sleep(200 * time.Millisecond)
|
|
|
|
// Verify that the output is still out of sync since there's a deferred
|
|
// update pending.
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
for _, chk := range checks.HealthChecks {
|
|
switch chk.CheckID {
|
|
case "web":
|
|
if chk.Output != "changed" {
|
|
t.Fatalf("unexpected update: %v", chk)
|
|
}
|
|
}
|
|
}
|
|
// Wait for the deferred update.
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Health.NodeChecks", &req, &checks); err != nil {
|
|
r.Fatal(err)
|
|
}
|
|
|
|
// Verify updated
|
|
for _, chk := range checks.HealthChecks {
|
|
switch chk.CheckID {
|
|
case "web":
|
|
if chk.Output != "deferred" {
|
|
r.Fatalf("no update: %v", chk)
|
|
}
|
|
}
|
|
}
|
|
})
|
|
|
|
}
|
|
|
|
func TestAgentAntiEntropy_NodeInfo(t *testing.T) {
|
|
t.Parallel()
|
|
nodeID := types.NodeID("40e4a748-2192-161a-0510-9bf59fe950b5")
|
|
nodeMeta := map[string]string{
|
|
"somekey": "somevalue",
|
|
}
|
|
a := &TestAgent{Name: t.Name(), HCL: `
|
|
node_id = "40e4a748-2192-161a-0510-9bf59fe950b5"
|
|
node_meta {
|
|
somekey = "somevalue"
|
|
}
|
|
`, NoInitialSync: true}
|
|
a.Start()
|
|
defer a.Shutdown()
|
|
|
|
// Register info
|
|
args := &structs.RegisterRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
Address: "127.0.0.1",
|
|
}
|
|
var out struct{}
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
|
|
req := structs.NodeSpecificRequest{
|
|
Datacenter: "dc1",
|
|
Node: a.Config.NodeName,
|
|
}
|
|
var services structs.IndexedNodeServices
|
|
// Wait for the sync
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Make sure we synced our node info - this should have ridden on the
|
|
// "consul" service sync
|
|
id := services.NodeServices.Node.ID
|
|
addrs := services.NodeServices.Node.TaggedAddresses
|
|
meta := services.NodeServices.Node.Meta
|
|
delete(meta, structs.MetaSegmentKey) // Added later, not in config.
|
|
if id != nodeID ||
|
|
!reflect.DeepEqual(addrs, a.config.TaggedAddresses) ||
|
|
!reflect.DeepEqual(meta, nodeMeta) {
|
|
r.Fatalf("bad: %v", services.NodeServices.Node)
|
|
}
|
|
})
|
|
|
|
// Blow away the catalog version of the node info
|
|
if err := a.RPC("Catalog.Register", args, &out); err != nil {
|
|
t.Fatalf("err: %v", err)
|
|
}
|
|
|
|
// Trigger anti-entropy run and wait
|
|
a.StartSync()
|
|
// Wait for the sync - this should have been a sync of just the node info
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Catalog.NodeServices", &req, &services); err != nil {
|
|
r.Fatalf("err: %v", err)
|
|
}
|
|
|
|
id := services.NodeServices.Node.ID
|
|
addrs := services.NodeServices.Node.TaggedAddresses
|
|
meta := services.NodeServices.Node.Meta
|
|
delete(meta, structs.MetaSegmentKey) // Added later, not in config.
|
|
if id != nodeID ||
|
|
!reflect.DeepEqual(addrs, a.config.TaggedAddresses) ||
|
|
!reflect.DeepEqual(meta, nodeMeta) {
|
|
r.Fatalf("bad: %v", services.NodeServices.Node)
|
|
}
|
|
})
|
|
}
|
|
|
|
func TestAgentAntiEntropy_deleteService_fails(t *testing.T) {
|
|
t.Parallel()
|
|
l := new(localState)
|
|
|
|
// todo(fs): data race
|
|
l.Lock()
|
|
defer l.Unlock()
|
|
if err := l.deleteService(""); err == nil {
|
|
t.Fatalf("should have failed")
|
|
}
|
|
}
|
|
|
|
func TestAgentAntiEntropy_deleteCheck_fails(t *testing.T) {
|
|
t.Parallel()
|
|
l := new(localState)
|
|
|
|
// todo(fs): data race
|
|
l.Lock()
|
|
defer l.Unlock()
|
|
if err := l.deleteCheck(""); err == nil {
|
|
t.Fatalf("should have errored")
|
|
}
|
|
}
|
|
|
|
func TestAgent_serviceTokens(t *testing.T) {
|
|
t.Parallel()
|
|
|
|
tokens := new(token.Store)
|
|
tokens.UpdateUserToken("default")
|
|
l := NewLocalState(config.DefaultRuntimeConfig(`bind_addr = "127.0.0.1" data_dir = "dummy"`), nil, tokens)
|
|
|
|
l.AddService(&structs.NodeService{
|
|
ID: "redis",
|
|
}, "")
|
|
|
|
// Returns default when no token is set
|
|
if token := l.ServiceToken("redis"); token != "default" {
|
|
t.Fatalf("bad: %s", token)
|
|
}
|
|
|
|
// Returns configured token
|
|
l.serviceTokens["redis"] = "abc123"
|
|
if token := l.ServiceToken("redis"); token != "abc123" {
|
|
t.Fatalf("bad: %s", token)
|
|
}
|
|
|
|
// Keeps token around for the delete
|
|
l.RemoveService("redis")
|
|
if token := l.ServiceToken("redis"); token != "abc123" {
|
|
t.Fatalf("bad: %s", token)
|
|
}
|
|
}
|
|
|
|
func TestAgent_checkTokens(t *testing.T) {
|
|
t.Parallel()
|
|
|
|
tokens := new(token.Store)
|
|
tokens.UpdateUserToken("default")
|
|
l := NewLocalState(config.DefaultRuntimeConfig(`bind_addr = "127.0.0.1" data_dir = "dummy"`), nil, tokens)
|
|
|
|
// Returns default when no token is set
|
|
if token := l.CheckToken("mem"); token != "default" {
|
|
t.Fatalf("bad: %s", token)
|
|
}
|
|
|
|
// Returns configured token
|
|
l.checkTokens["mem"] = "abc123"
|
|
if token := l.CheckToken("mem"); token != "abc123" {
|
|
t.Fatalf("bad: %s", token)
|
|
}
|
|
|
|
// Keeps token around for the delete
|
|
l.RemoveCheck("mem")
|
|
if token := l.CheckToken("mem"); token != "abc123" {
|
|
t.Fatalf("bad: %s", token)
|
|
}
|
|
}
|
|
|
|
func TestAgent_checkCriticalTime(t *testing.T) {
|
|
t.Parallel()
|
|
l := NewLocalState(config.DefaultRuntimeConfig(`bind_addr = "127.0.0.1" data_dir = "dummy"`), nil, new(token.Store))
|
|
|
|
svc := &structs.NodeService{ID: "redis", Service: "redis", Port: 8000}
|
|
l.AddService(svc, "")
|
|
|
|
// Add a passing check and make sure it's not critical.
|
|
checkID := types.CheckID("redis:1")
|
|
chk := &structs.HealthCheck{
|
|
Node: "node",
|
|
CheckID: checkID,
|
|
Name: "redis:1",
|
|
ServiceID: "redis",
|
|
Status: api.HealthPassing,
|
|
}
|
|
l.AddCheck(chk, "")
|
|
if checks := l.CriticalChecks(); len(checks) > 0 {
|
|
t.Fatalf("should not have any critical checks")
|
|
}
|
|
|
|
// Set it to warning and make sure that doesn't show up as critical.
|
|
l.UpdateCheck(checkID, api.HealthWarning, "")
|
|
if checks := l.CriticalChecks(); len(checks) > 0 {
|
|
t.Fatalf("should not have any critical checks")
|
|
}
|
|
|
|
// Fail the check and make sure the time looks reasonable.
|
|
l.UpdateCheck(checkID, api.HealthCritical, "")
|
|
if crit, ok := l.CriticalChecks()[checkID]; !ok {
|
|
t.Fatalf("should have a critical check")
|
|
} else if crit.CriticalFor > time.Millisecond {
|
|
t.Fatalf("bad: %#v", crit)
|
|
}
|
|
|
|
// Wait a while, then fail it again and make sure the time keeps track
|
|
// of the initial failure, and doesn't reset here.
|
|
time.Sleep(50 * time.Millisecond)
|
|
l.UpdateCheck(chk.CheckID, api.HealthCritical, "")
|
|
if crit, ok := l.CriticalChecks()[checkID]; !ok {
|
|
t.Fatalf("should have a critical check")
|
|
} else if crit.CriticalFor < 25*time.Millisecond ||
|
|
crit.CriticalFor > 75*time.Millisecond {
|
|
t.Fatalf("bad: %#v", crit)
|
|
}
|
|
|
|
// Set it passing again.
|
|
l.UpdateCheck(checkID, api.HealthPassing, "")
|
|
if checks := l.CriticalChecks(); len(checks) > 0 {
|
|
t.Fatalf("should not have any critical checks")
|
|
}
|
|
|
|
// Fail the check and make sure the time looks like it started again
|
|
// from the latest failure, not the original one.
|
|
l.UpdateCheck(checkID, api.HealthCritical, "")
|
|
if crit, ok := l.CriticalChecks()[checkID]; !ok {
|
|
t.Fatalf("should have a critical check")
|
|
} else if crit.CriticalFor > time.Millisecond {
|
|
t.Fatalf("bad: %#v", crit)
|
|
}
|
|
}
|
|
|
|
func TestAgent_AddCheckFailure(t *testing.T) {
|
|
t.Parallel()
|
|
l := NewLocalState(config.DefaultRuntimeConfig(`bind_addr = "127.0.0.1" data_dir = "dummy"`), nil, new(token.Store))
|
|
|
|
// Add a check for a service that does not exist and verify that it fails
|
|
checkID := types.CheckID("redis:1")
|
|
chk := &structs.HealthCheck{
|
|
Node: "node",
|
|
CheckID: checkID,
|
|
Name: "redis:1",
|
|
ServiceID: "redis",
|
|
Status: api.HealthPassing,
|
|
}
|
|
expectedErr := "ServiceID \"redis\" does not exist"
|
|
if err := l.AddCheck(chk, ""); err == nil || expectedErr != err.Error() {
|
|
t.Fatalf("Expected error when adding a check for a non-existent service but got %v", err)
|
|
}
|
|
|
|
}
|
|
|
|
func TestAgent_nestedPauseResume(t *testing.T) {
|
|
t.Parallel()
|
|
l := new(localState)
|
|
if l.isPaused() != false {
|
|
t.Fatal("localState should be unPaused after init")
|
|
}
|
|
l.Pause()
|
|
if l.isPaused() != true {
|
|
t.Fatal("localState should be Paused after first call to Pause()")
|
|
}
|
|
l.Pause()
|
|
if l.isPaused() != true {
|
|
t.Fatal("localState should STILL be Paused after second call to Pause()")
|
|
}
|
|
l.Resume()
|
|
if l.isPaused() != true {
|
|
t.Fatal("localState should STILL be Paused after FIRST call to Resume()")
|
|
}
|
|
l.Resume()
|
|
if l.isPaused() != false {
|
|
t.Fatal("localState should NOT be Paused after SECOND call to Resume()")
|
|
}
|
|
|
|
defer func() {
|
|
err := recover()
|
|
if err == nil {
|
|
t.Fatal("unbalanced Resume() should cause a panic()")
|
|
}
|
|
}()
|
|
l.Resume()
|
|
}
|
|
|
|
func TestAgent_sendCoordinate(t *testing.T) {
|
|
t.Parallel()
|
|
a := NewTestAgent(t.Name(), `
|
|
sync_coordinate_interval_min = "1ms"
|
|
sync_coordinate_rate_target = 10.0
|
|
consul = {
|
|
coordinate = {
|
|
update_period = "100ms"
|
|
update_batch_size = 10
|
|
update_max_batches = 1
|
|
}
|
|
}
|
|
`)
|
|
defer a.Shutdown()
|
|
|
|
t.Logf("%d %d %s", a.consulConfig().CoordinateUpdateBatchSize, a.consulConfig().CoordinateUpdateMaxBatches,
|
|
a.consulConfig().CoordinateUpdatePeriod.String())
|
|
|
|
// Make sure the coordinate is present.
|
|
req := structs.DCSpecificRequest{
|
|
Datacenter: a.Config.Datacenter,
|
|
}
|
|
var reply structs.IndexedCoordinates
|
|
retry.Run(t, func(r *retry.R) {
|
|
if err := a.RPC("Coordinate.ListNodes", &req, &reply); err != nil {
|
|
r.Fatalf("err: %s", err)
|
|
}
|
|
if len(reply.Coordinates) != 1 {
|
|
r.Fatalf("expected a coordinate: %v", reply)
|
|
}
|
|
coord := reply.Coordinates[0]
|
|
if coord.Node != a.Config.NodeName || coord.Coord == nil {
|
|
r.Fatalf("bad: %v", coord)
|
|
}
|
|
})
|
|
}
|