b656981cf0
Plan rejections occur when the scheduler work and the leader plan applier disagree on the feasibility of a plan. This may happen for valid reasons: since Nomad does parallel scheduling, it is expected that different workers will have a different state when computing placements. As the final plan reaches the leader plan applier, it may no longer be valid due to a concurrent scheduling taking up intended resources. In these situations the plan applier will notify the worker that the plan was rejected and that they should refresh their state before trying again. In some rare and unexpected circumstances it has been observed that workers will repeatedly submit the same plan, even if they are always rejected. While the root cause is still unknown this mitigation has been put in place. The plan applier will now track the history of plan rejections per client and include in the plan result a list of node IDs that should be set as ineligible if the number of rejections in a given time window crosses a certain threshold. The window size and threshold value can be adjusted in the server configuration. To avoid marking several nodes as ineligible at one, the operation is rate limited to 5 nodes every 30min, with an initial burst of 10 operations.
271 lines
8.9 KiB
Go
271 lines
8.9 KiB
Go
package agent
|
|
|
|
import (
|
|
"bytes"
|
|
"fmt"
|
|
"io"
|
|
"os"
|
|
"path/filepath"
|
|
"time"
|
|
|
|
"github.com/hashicorp/hcl"
|
|
client "github.com/hashicorp/nomad/client/config"
|
|
"github.com/hashicorp/nomad/helper"
|
|
"github.com/hashicorp/nomad/nomad/structs/config"
|
|
)
|
|
|
|
// ParseConfigFile returns an agent.Config from parsed from a file.
|
|
func ParseConfigFile(path string) (*Config, error) {
|
|
// slurp
|
|
var buf bytes.Buffer
|
|
path, err := filepath.Abs(path)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
f, err := os.Open(path)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
defer f.Close()
|
|
if _, err := io.Copy(&buf, f); err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
// parse
|
|
c := &Config{
|
|
Client: &ClientConfig{
|
|
ServerJoin: &ServerJoin{},
|
|
TemplateConfig: &client.ClientTemplateConfig{
|
|
Wait: &client.WaitConfig{},
|
|
WaitBounds: &client.WaitConfig{},
|
|
ConsulRetry: &client.RetryConfig{},
|
|
VaultRetry: &client.RetryConfig{},
|
|
},
|
|
},
|
|
Server: &ServerConfig{
|
|
PlanRejectionTracker: &PlanRejectionTracker{},
|
|
ServerJoin: &ServerJoin{},
|
|
},
|
|
ACL: &ACLConfig{},
|
|
Audit: &config.AuditConfig{},
|
|
Consul: &config.ConsulConfig{},
|
|
Autopilot: &config.AutopilotConfig{},
|
|
Telemetry: &Telemetry{},
|
|
Vault: &config.VaultConfig{},
|
|
}
|
|
|
|
err = hcl.Decode(c, buf.String())
|
|
if err != nil {
|
|
return nil, fmt.Errorf("failed to decode HCL file %s: %w", path, err)
|
|
}
|
|
|
|
// convert strings to time.Durations
|
|
tds := []durationConversionMap{
|
|
{"gc_interval", &c.Client.GCInterval, &c.Client.GCIntervalHCL, nil},
|
|
{"acl.token_ttl", &c.ACL.TokenTTL, &c.ACL.TokenTTLHCL, nil},
|
|
{"acl.policy_ttl", &c.ACL.PolicyTTL, &c.ACL.PolicyTTLHCL, nil},
|
|
{"client.server_join.retry_interval", &c.Client.ServerJoin.RetryInterval, &c.Client.ServerJoin.RetryIntervalHCL, nil},
|
|
{"server.heartbeat_grace", &c.Server.HeartbeatGrace, &c.Server.HeartbeatGraceHCL, nil},
|
|
{"server.min_heartbeat_ttl", &c.Server.MinHeartbeatTTL, &c.Server.MinHeartbeatTTLHCL, nil},
|
|
{"server.failover_heartbeat_ttl", &c.Server.FailoverHeartbeatTTL, &c.Server.FailoverHeartbeatTTLHCL, nil},
|
|
{"server.plan_rejection_tracker.node_window", &c.Server.PlanRejectionTracker.NodeWindow, &c.Server.PlanRejectionTracker.NodeWindowHCL, nil},
|
|
{"server.retry_interval", &c.Server.RetryInterval, &c.Server.RetryIntervalHCL, nil},
|
|
{"server.server_join.retry_interval", &c.Server.ServerJoin.RetryInterval, &c.Server.ServerJoin.RetryIntervalHCL, nil},
|
|
{"consul.timeout", &c.Consul.Timeout, &c.Consul.TimeoutHCL, nil},
|
|
{"autopilot.server_stabilization_time", &c.Autopilot.ServerStabilizationTime, &c.Autopilot.ServerStabilizationTimeHCL, nil},
|
|
{"autopilot.last_contact_threshold", &c.Autopilot.LastContactThreshold, &c.Autopilot.LastContactThresholdHCL, nil},
|
|
{"telemetry.collection_interval", &c.Telemetry.collectionInterval, &c.Telemetry.CollectionInterval, nil},
|
|
{"client.template.block_query_wait", nil, &c.Client.TemplateConfig.BlockQueryWaitTimeHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.BlockQueryWaitTime = d
|
|
},
|
|
},
|
|
{"client.template.max_stale", nil, &c.Client.TemplateConfig.MaxStaleHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.MaxStale = d
|
|
}},
|
|
{"client.template.wait.min", nil, &c.Client.TemplateConfig.Wait.MinHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.Wait.Min = d
|
|
},
|
|
},
|
|
{"client.template.wait.max", nil, &c.Client.TemplateConfig.Wait.MaxHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.Wait.Max = d
|
|
},
|
|
},
|
|
{"client.template.wait_bounds.min", nil, &c.Client.TemplateConfig.WaitBounds.MinHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.WaitBounds.Min = d
|
|
},
|
|
},
|
|
{"client.template.wait_bounds.max", nil, &c.Client.TemplateConfig.WaitBounds.MaxHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.WaitBounds.Max = d
|
|
},
|
|
},
|
|
{"client.template.consul_retry.backoff", nil, &c.Client.TemplateConfig.ConsulRetry.BackoffHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.ConsulRetry.Backoff = d
|
|
},
|
|
},
|
|
{"client.template.consul_retry.max_backoff", nil, &c.Client.TemplateConfig.ConsulRetry.MaxBackoffHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.ConsulRetry.MaxBackoff = d
|
|
},
|
|
},
|
|
{"client.template.vault_retry.backoff", nil, &c.Client.TemplateConfig.VaultRetry.BackoffHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.VaultRetry.Backoff = d
|
|
},
|
|
},
|
|
{"client.template.vault_retry.max_backoff", nil, &c.Client.TemplateConfig.VaultRetry.MaxBackoffHCL,
|
|
func(d *time.Duration) {
|
|
c.Client.TemplateConfig.VaultRetry.MaxBackoff = d
|
|
},
|
|
},
|
|
}
|
|
|
|
// Add enterprise audit sinks for time.Duration parsing
|
|
for i, sink := range c.Audit.Sinks {
|
|
tds = append(tds, durationConversionMap{
|
|
fmt.Sprintf("audit.sink.%d", i), &sink.RotateDuration, &sink.RotateDurationHCL, nil})
|
|
}
|
|
|
|
// convert strings to time.Durations
|
|
err = convertDurations(tds)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
// report unexpected keys
|
|
err = extraKeys(c)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
// Set client template config or its members to nil if not set.
|
|
finalizeClientTemplateConfig(c)
|
|
|
|
return c, nil
|
|
}
|
|
|
|
// durationConversionMap holds args for one duration conversion
|
|
type durationConversionMap struct {
|
|
targetFieldPath string
|
|
targetField *time.Duration
|
|
sourceField *string
|
|
setFunc func(*time.Duration)
|
|
}
|
|
|
|
// convertDurations parses the duration strings specified in the config files
|
|
// into time.Durations
|
|
func convertDurations(xs []durationConversionMap) error {
|
|
for _, x := range xs {
|
|
// if targetField is not a pointer itself, use the field map.
|
|
if x.targetField != nil && x.sourceField != nil && "" != *x.sourceField {
|
|
d, err := time.ParseDuration(*x.sourceField)
|
|
if err != nil {
|
|
return fmt.Errorf("%s can't parse time duration %s", x.targetFieldPath, *x.sourceField)
|
|
}
|
|
|
|
*x.targetField = d
|
|
} else if x.setFunc != nil && x.sourceField != nil && "" != *x.sourceField {
|
|
// if targetField is a pointer itself, use the setFunc closure.
|
|
d, err := time.ParseDuration(*x.sourceField)
|
|
if err != nil {
|
|
return fmt.Errorf("%s can't parse time duration %s", x.targetFieldPath, *x.sourceField)
|
|
}
|
|
x.setFunc(&d)
|
|
}
|
|
}
|
|
|
|
return nil
|
|
}
|
|
|
|
func extraKeys(c *Config) error {
|
|
// hcl leaves behind extra keys when parsing JSON. These keys
|
|
// are kept on the top level, taken from slices or the keys of
|
|
// structs contained in slices. Clean up before looking for
|
|
// extra keys.
|
|
for range c.HTTPAPIResponseHeaders {
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, "http_api_response_headers")
|
|
}
|
|
|
|
for _, p := range c.Plugins {
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, p.Name)
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, "config")
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, "plugin")
|
|
}
|
|
|
|
for _, k := range []string{"options", "meta", "chroot_env", "servers", "server_join"} {
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, k)
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, "client")
|
|
}
|
|
|
|
// stats is an unused key, continue to silently ignore it
|
|
helper.RemoveEqualFold(&c.Client.ExtraKeysHCL, "stats")
|
|
|
|
// Remove HostVolume extra keys
|
|
for _, hv := range c.Client.HostVolumes {
|
|
helper.RemoveEqualFold(&c.Client.ExtraKeysHCL, hv.Name)
|
|
helper.RemoveEqualFold(&c.Client.ExtraKeysHCL, "host_volume")
|
|
}
|
|
|
|
// Remove HostNetwork extra keys
|
|
for _, hn := range c.Client.HostNetworks {
|
|
helper.RemoveEqualFold(&c.Client.ExtraKeysHCL, hn.Name)
|
|
helper.RemoveEqualFold(&c.Client.ExtraKeysHCL, "host_network")
|
|
}
|
|
|
|
// Remove AuditConfig extra keys
|
|
for _, f := range c.Audit.Filters {
|
|
helper.RemoveEqualFold(&c.Audit.ExtraKeysHCL, f.Name)
|
|
helper.RemoveEqualFold(&c.Audit.ExtraKeysHCL, "filter")
|
|
}
|
|
|
|
for _, s := range c.Audit.Sinks {
|
|
helper.RemoveEqualFold(&c.Audit.ExtraKeysHCL, s.Name)
|
|
helper.RemoveEqualFold(&c.Audit.ExtraKeysHCL, "sink")
|
|
}
|
|
|
|
for _, k := range []string{"enabled_schedulers", "start_join", "retry_join", "server_join"} {
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, k)
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, "server")
|
|
}
|
|
|
|
for _, k := range []string{"datadog_tags"} {
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, k)
|
|
helper.RemoveEqualFold(&c.ExtraKeysHCL, "telemetry")
|
|
}
|
|
|
|
return helper.UnusedKeys(c)
|
|
}
|
|
|
|
// hcl.Decode will error if the ClientTemplateConfig isn't initialized with empty
|
|
// structs, however downstream code expect nils if the struct only contains fields
|
|
// with the zero value for its type. This function nils out type members that are
|
|
// structs where all the member fields are just the zero value for its type.
|
|
func finalizeClientTemplateConfig(config *Config) {
|
|
if config.Client.TemplateConfig.Wait.IsEmpty() {
|
|
config.Client.TemplateConfig.Wait = nil
|
|
}
|
|
|
|
if config.Client.TemplateConfig.WaitBounds.IsEmpty() {
|
|
config.Client.TemplateConfig.WaitBounds = nil
|
|
}
|
|
|
|
if config.Client.TemplateConfig.ConsulRetry.IsEmpty() {
|
|
config.Client.TemplateConfig.ConsulRetry = nil
|
|
}
|
|
|
|
if config.Client.TemplateConfig.VaultRetry.IsEmpty() {
|
|
config.Client.TemplateConfig.VaultRetry = nil
|
|
}
|
|
|
|
if config.Client.TemplateConfig.IsEmpty() {
|
|
config.Client.TemplateConfig = nil
|
|
}
|
|
}
|