open-vault/command/server.go

2634 lines
74 KiB
Go
Raw Normal View History

2015-03-12 22:21:11 +00:00
package command
import (
"context"
"crypto/sha256"
"encoding/base64"
"encoding/hex"
2015-03-12 22:30:07 +00:00
"fmt"
"io"
"io/ioutil"
2015-03-13 17:09:38 +00:00
"net"
"net/http"
"net/url"
2015-04-04 19:06:41 +00:00
"os"
"path/filepath"
2015-10-28 17:05:56 +00:00
"runtime"
2015-04-04 19:06:41 +00:00
"sort"
"strconv"
2015-03-12 22:21:11 +00:00
"strings"
"sync"
"time"
2015-03-12 22:21:11 +00:00
monitoring "cloud.google.com/go/monitoring/apiv3"
"github.com/armon/go-metrics"
"github.com/armon/go-metrics/circonus"
"github.com/armon/go-metrics/datadog"
"github.com/armon/go-metrics/prometheus"
stackdriver "github.com/google/go-metrics-stackdriver"
2016-04-04 14:44:22 +00:00
"github.com/hashicorp/errwrap"
log "github.com/hashicorp/go-hclog"
wrapping "github.com/hashicorp/go-kms-wrapping"
aeadwrapper "github.com/hashicorp/go-kms-wrapping/wrappers/aead"
"github.com/hashicorp/go-multierror"
"github.com/hashicorp/go-sockaddr"
2015-04-05 01:07:53 +00:00
"github.com/hashicorp/vault/audit"
2015-03-12 22:30:07 +00:00
"github.com/hashicorp/vault/command/server"
2018-09-18 03:03:00 +00:00
serverseal "github.com/hashicorp/vault/command/server/seal"
2018-11-07 01:21:24 +00:00
"github.com/hashicorp/vault/helper/builtinplugins"
"github.com/hashicorp/vault/helper/metricsutil"
"github.com/hashicorp/vault/helper/namespace"
"github.com/hashicorp/vault/helper/reload"
vaulthttp "github.com/hashicorp/vault/http"
2020-01-23 19:00:49 +00:00
"github.com/hashicorp/vault/sdk/helper/gatedwriter"
"github.com/hashicorp/vault/sdk/helper/jsonutil"
"github.com/hashicorp/vault/sdk/helper/logging"
"github.com/hashicorp/vault/sdk/helper/mlock"
"github.com/hashicorp/vault/sdk/helper/parseutil"
"github.com/hashicorp/vault/sdk/helper/useragent"
"github.com/hashicorp/vault/sdk/logical"
"github.com/hashicorp/vault/sdk/physical"
"github.com/hashicorp/vault/sdk/version"
sr "github.com/hashicorp/vault/serviceregistration"
2015-03-13 17:09:38 +00:00
"github.com/hashicorp/vault/vault"
vaultseal "github.com/hashicorp/vault/vault/seal"
"github.com/mitchellh/cli"
"github.com/mitchellh/go-testing-interface"
"github.com/posener/complete"
2019-11-07 16:54:34 +00:00
"go.uber.org/atomic"
"golang.org/x/net/http/httpproxy"
"google.golang.org/api/option"
"google.golang.org/grpc/grpclog"
2015-03-12 22:21:11 +00:00
)
2017-09-08 02:03:44 +00:00
var _ cli.Command = (*ServerCommand)(nil)
var _ cli.CommandAutocomplete = (*ServerCommand)(nil)
var memProfilerEnabled = false
var enableFourClusterDev = func(c *ServerCommand, base *vault.CoreConfig, info map[string]string, infoKeys []string, devListenAddress, tempDir string) int {
c.logger.Error("-dev-four-cluster only supported in enterprise Vault")
return 1
}
2018-10-23 06:34:02 +00:00
const storageMigrationLock = "core/migration"
2015-03-12 22:21:11 +00:00
type ServerCommand struct {
*BaseCommand
2015-04-05 01:07:53 +00:00
AuditBackends map[string]audit.Factory
CredentialBackends map[string]logical.Factory
LogicalBackends map[string]logical.Factory
PhysicalBackends map[string]physical.Factory
2015-03-20 18:32:18 +00:00
ServiceRegistrations map[string]sr.Factory
ShutdownCh chan struct{}
SighupCh chan struct{}
SigUSR2Ch chan struct{}
WaitGroup *sync.WaitGroup
logWriter io.Writer
logGate *gatedwriter.Writer
logger log.Logger
2016-06-02 16:40:25 +00:00
cleanupGuard sync.Once
2016-09-30 04:06:40 +00:00
reloadFuncsLock *sync.RWMutex
reloadFuncs *map[string][]reload.ReloadFunc
2017-09-08 02:03:44 +00:00
startedCh chan (struct{}) // for tests
reloadedCh chan (struct{}) // for tests
// new stuff
2019-07-24 16:41:28 +00:00
flagConfigs []string
flagLogLevel string
flagLogFormat string
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
flagRecovery bool
2019-07-24 16:41:28 +00:00
flagDev bool
flagDevRootTokenID string
flagDevListenAddr string
flagDevNoStoreToken bool
2017-09-08 02:03:44 +00:00
flagDevPluginDir string
flagDevPluginInit bool
2017-09-08 02:03:44 +00:00
flagDevHA bool
flagDevLatency int
flagDevLatencyJitter int
2017-09-08 02:03:44 +00:00
flagDevLeasedKV bool
flagDevKVV1 bool
flagDevSkipInit bool
2017-09-08 02:03:44 +00:00
flagDevThreeNode bool
2018-02-22 05:23:37 +00:00
flagDevFourCluster bool
flagDevTransactional bool
flagDevAutoSeal bool
2017-09-08 02:03:44 +00:00
flagTestVerifyOnly bool
flagCombineLogs bool
flagTestServerConfig bool
flagDevConsul bool
2017-09-08 02:03:44 +00:00
}
2018-04-17 22:52:09 +00:00
type ServerListener struct {
net.Listener
config map[string]interface{}
maxRequestSize int64
maxRequestDuration time.Duration
unauthenticatedMetricsAccess bool
2018-04-17 22:52:09 +00:00
}
2017-09-08 02:03:44 +00:00
func (c *ServerCommand) Synopsis() string {
return "Start a Vault server"
}
func (c *ServerCommand) Help() string {
helpText := `
Usage: vault server [options]
This command starts a Vault server that responds to API requests. By default,
Vault will start in a "sealed" state. The Vault cluster must be initialized
before use, usually by the "vault operator init" command. Each Vault server must
also be unsealed using the "vault operator unseal" command or the API before the
server can respond to requests.
2017-09-08 02:03:44 +00:00
Start a server with a configuration file:
$ vault server -config=/etc/vault/config.hcl
Run in "dev" mode:
$ vault server -dev -dev-root-token-id="root"
For a full list of examples, please see the documentation.
` + c.Flags().Help()
return strings.TrimSpace(helpText)
}
func (c *ServerCommand) Flags() *FlagSets {
set := c.flagSet(FlagSetHTTP)
f := set.NewFlagSet("Command Options")
f.StringSliceVar(&StringSliceVar{
Name: "config",
Target: &c.flagConfigs,
Completion: complete.PredictOr(
complete.PredictFiles("*.hcl"),
complete.PredictFiles("*.json"),
complete.PredictDirs("*"),
),
Usage: "Path to a configuration file or directory of configuration " +
"files. This flag can be specified multiple times to load multiple " +
"configurations. If the path is a directory, all files which end in " +
".hcl or .json are loaded.",
})
f.StringVar(&StringVar{
Name: "log-level",
Target: &c.flagLogLevel,
Default: notSetValue,
EnvVar: "VAULT_LOG_LEVEL",
2017-09-08 02:03:44 +00:00
Completion: complete.PredictSet("trace", "debug", "info", "warn", "err"),
Usage: "Log verbosity level. Supported values (in order of detail) are " +
"\"trace\", \"debug\", \"info\", \"warn\", and \"err\".",
})
f.StringVar(&StringVar{
Name: "log-format",
Target: &c.flagLogFormat,
Default: notSetValue,
// EnvVar can't be just "VAULT_LOG_FORMAT", because more than one env var name is supported
// for backwards compatibility reasons.
// See github.com/hashicorp/vault/sdk/helper/logging.ParseEnvLogFormat()
Completion: complete.PredictSet("standard", "json"),
Usage: `Log format. Supported values are "standard" and "json".`,
})
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
f.BoolVar(&BoolVar{
Name: "recovery",
Target: &c.flagRecovery,
Usage: "Enable recovery mode. In this mode, Vault is used to perform recovery actions." +
"Using a recovery operation token, \"sys/raw\" API can be used to manipulate the storage.",
})
2017-09-08 02:03:44 +00:00
f = set.NewFlagSet("Dev Options")
f.BoolVar(&BoolVar{
Name: "dev",
Target: &c.flagDev,
Usage: "Enable development mode. In this mode, Vault runs in-memory and " +
"starts unsealed. As the name implies, do not run \"dev\" mode in " +
"production.",
})
f.StringVar(&StringVar{
Name: "dev-root-token-id",
Target: &c.flagDevRootTokenID,
Default: "",
EnvVar: "VAULT_DEV_ROOT_TOKEN_ID",
Usage: "Initial root token. This only applies when running in \"dev\" " +
"mode.",
})
f.StringVar(&StringVar{
Name: "dev-listen-address",
Target: &c.flagDevListenAddr,
Default: "127.0.0.1:8200",
EnvVar: "VAULT_DEV_LISTEN_ADDRESS",
Usage: "Address to bind to in \"dev\" mode.",
})
f.BoolVar(&BoolVar{
Name: "dev-no-store-token",
Target: &c.flagDevNoStoreToken,
Default: false,
Usage: "Do not persist the dev root token to the token helper " +
"(usually the local filesystem) for use in future requests. " +
"The token will only be displayed in the command output.",
})
2017-09-08 02:03:44 +00:00
// Internal-only flags to follow.
//
// Why hello there little source code reader! Welcome to the Vault source
// code. The remaining options are intentionally undocumented and come with
// no warranty or backwards-compatibility promise. Do not use these flags
2017-09-08 02:03:44 +00:00
// in production. Do not build automation using these flags. Unless you are
// developing against Vault, you should not need any of these flags.
f.StringVar(&StringVar{
Name: "dev-plugin-dir",
Target: &c.flagDevPluginDir,
Default: "",
Completion: complete.PredictDirs("*"),
Hidden: true,
})
f.BoolVar(&BoolVar{
Name: "dev-plugin-init",
Target: &c.flagDevPluginInit,
Default: true,
Hidden: true,
})
2017-09-08 02:03:44 +00:00
f.BoolVar(&BoolVar{
Name: "dev-ha",
Target: &c.flagDevHA,
Default: false,
Hidden: true,
})
f.BoolVar(&BoolVar{
Name: "dev-transactional",
Target: &c.flagDevTransactional,
Default: false,
Hidden: true,
})
f.IntVar(&IntVar{
Name: "dev-latency",
Target: &c.flagDevLatency,
Hidden: true,
})
f.IntVar(&IntVar{
Name: "dev-latency-jitter",
Target: &c.flagDevLatencyJitter,
Hidden: true,
})
2017-09-08 02:03:44 +00:00
f.BoolVar(&BoolVar{
Name: "dev-leased-kv",
2017-09-08 02:03:44 +00:00
Target: &c.flagDevLeasedKV,
Default: false,
Hidden: true,
})
f.BoolVar(&BoolVar{
Name: "dev-kv-v1",
Target: &c.flagDevKVV1,
Default: false,
Hidden: true,
})
f.BoolVar(&BoolVar{
Name: "dev-auto-seal",
Target: &c.flagDevAutoSeal,
Default: false,
Hidden: true,
})
f.BoolVar(&BoolVar{
Name: "dev-skip-init",
Target: &c.flagDevSkipInit,
Default: false,
Hidden: true,
})
2017-09-08 02:03:44 +00:00
f.BoolVar(&BoolVar{
Name: "dev-three-node",
Target: &c.flagDevThreeNode,
Default: false,
Hidden: true,
})
2018-02-22 05:23:37 +00:00
f.BoolVar(&BoolVar{
Name: "dev-four-cluster",
Target: &c.flagDevFourCluster,
Default: false,
Hidden: true,
})
f.BoolVar(&BoolVar{
Name: "dev-consul",
Target: &c.flagDevConsul,
Default: false,
Hidden: true,
})
// TODO: should the below flags be public?
f.BoolVar(&BoolVar{
Name: "combine-logs",
Target: &c.flagCombineLogs,
Default: false,
Hidden: true,
})
2017-09-08 02:03:44 +00:00
f.BoolVar(&BoolVar{
Name: "test-verify-only",
Target: &c.flagTestVerifyOnly,
Default: false,
Hidden: true,
})
f.BoolVar(&BoolVar{
Name: "test-server-config",
Target: &c.flagTestServerConfig,
Default: false,
Hidden: true,
})
2017-09-08 02:03:44 +00:00
// End internal-only flags.
return set
}
func (c *ServerCommand) AutocompleteArgs() complete.Predictor {
return complete.PredictNothing
}
func (c *ServerCommand) AutocompleteFlags() complete.Flags {
return c.Flags().Completions()
2015-03-12 22:21:11 +00:00
}
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
func (c *ServerCommand) parseConfig() (*server.Config, error) {
2015-03-12 22:21:11 +00:00
// Load the configuration
2015-03-12 22:30:07 +00:00
var config *server.Config
for _, path := range c.flagConfigs {
current, err := server.LoadConfig(path)
2015-03-12 22:30:07 +00:00
if err != nil {
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
return nil, errwrap.Wrapf(fmt.Sprintf("error loading configuration from %s: {{err}}", path), err)
2015-03-12 22:30:07 +00:00
}
if config == nil {
config = current
} else {
config = config.Merge(current)
}
}
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
return config, nil
}
func (c *ServerCommand) runRecoveryMode() int {
config, err := c.parseConfig()
if err != nil {
c.UI.Error(err.Error())
return 1
}
2015-03-12 22:30:07 +00:00
// Ensure at least one config was found.
if config == nil {
c.UI.Output(wrapAtLength(
"No configuration files found. Please provide configurations with the " +
"-config flag. If you are supplying the path to a directory, please " +
"ensure the directory contains files with the .hcl or .json " +
"extension."))
return 1
}
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
level, logLevelString, logLevelWasNotSet, logFormat, err := c.processLogLevelAndFormat(config)
if err != nil {
c.UI.Error(err.Error())
return 1
}
c.logger = log.New(&log.LoggerOptions{
Output: c.logWriter,
Level: level,
// Note that if logFormat is either unspecified or standard, then
// the resulting logger's format will be standard.
JSONFormat: logFormat == logging.JSONFormat,
})
logLevelStr, err := c.adjustLogLevel(config, logLevelWasNotSet)
if err != nil {
c.UI.Error(err.Error())
return 1
}
if logLevelStr != "" {
logLevelString = logLevelStr
}
// create GRPC logger
namedGRPCLogFaker := c.logger.Named("grpclogfaker")
grpclog.SetLogger(&grpclogFaker{
logger: namedGRPCLogFaker,
log: os.Getenv("VAULT_GRPC_LOGGING") != "",
})
if config.Storage == nil {
c.UI.Output("A storage backend must be specified")
return 1
}
if config.DefaultMaxRequestDuration != 0 {
vault.DefaultMaxRequestDuration = config.DefaultMaxRequestDuration
}
proxyCfg := httpproxy.FromEnvironment()
c.logger.Info("proxy environment", "http_proxy", proxyCfg.HTTPProxy,
"https_proxy", proxyCfg.HTTPSProxy, "no_proxy", proxyCfg.NoProxy)
// Initialize the storage backend
factory, exists := c.PhysicalBackends[config.Storage.Type]
if !exists {
c.UI.Error(fmt.Sprintf("Unknown storage type %s", config.Storage.Type))
return 1
}
if config.Storage.Type == "raft" {
if envCA := os.Getenv("VAULT_CLUSTER_ADDR"); envCA != "" {
config.ClusterAddr = envCA
}
if len(config.ClusterAddr) == 0 {
c.UI.Error("Cluster address must be set when using raft storage")
return 1
}
}
namedStorageLogger := c.logger.Named("storage." + config.Storage.Type)
backend, err := factory(config.Storage.Config, namedStorageLogger)
if err != nil {
c.UI.Error(fmt.Sprintf("Error initializing storage of type %s: %s", config.Storage.Type, err))
return 1
}
infoKeys := make([]string, 0, 10)
info := make(map[string]string)
info["log level"] = logLevelString
infoKeys = append(infoKeys, "log level")
var barrierSeal vault.Seal
var sealConfigError error
if len(config.Seals) == 0 {
config.Seals = append(config.Seals, &server.Seal{Type: wrapping.Shamir})
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
}
if len(config.Seals) > 1 {
c.UI.Error("Only one seal block is accepted in recovery mode")
return 1
}
configSeal := config.Seals[0]
sealType := wrapping.Shamir
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
if !configSeal.Disabled && os.Getenv("VAULT_SEAL_TYPE") != "" {
sealType = os.Getenv("VAULT_SEAL_TYPE")
configSeal.Type = sealType
} else {
sealType = configSeal.Type
}
var seal vault.Seal
sealLogger := c.logger.Named(sealType)
seal, sealConfigError = serverseal.ConfigureSeal(configSeal, &infoKeys, &info, sealLogger, vault.NewDefaultSeal(&vaultseal.Access{
Wrapper: aeadwrapper.NewWrapper(&wrapping.WrapperOptions{
Logger: c.logger.Named("shamir"),
}),
}))
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
if sealConfigError != nil {
if !errwrap.ContainsType(sealConfigError, new(logical.KeyNotFoundError)) {
c.UI.Error(fmt.Sprintf(
"Error parsing Seal configuration: %s", sealConfigError))
return 1
}
}
if seal == nil {
c.UI.Error(fmt.Sprintf(
"After configuring seal nil returned, seal type was %s", sealType))
return 1
}
barrierSeal = seal
// Ensure that the seal finalizer is called, even if using verify-only
defer func() {
err = seal.Finalize(context.Background())
if err != nil {
c.UI.Error(fmt.Sprintf("Error finalizing seals: %v", err))
}
}()
coreConfig := &vault.CoreConfig{
Physical: backend,
StorageType: config.Storage.Type,
Seal: barrierSeal,
Logger: c.logger,
DisableMlock: config.DisableMlock,
RecoveryMode: c.flagRecovery,
ClusterAddr: config.ClusterAddr,
}
core, newCoreError := vault.NewCore(coreConfig)
if newCoreError != nil {
if vault.IsFatalError(newCoreError) {
c.UI.Error(fmt.Sprintf("Error initializing core: %s", newCoreError))
return 1
}
}
if err := core.InitializeRecovery(context.Background()); err != nil {
c.UI.Error(fmt.Sprintf("Error initializing core in recovery mode: %s", err))
return 1
}
// Compile server information for output later
infoKeys = append(infoKeys, "storage")
info["storage"] = config.Storage.Type
if coreConfig.ClusterAddr != "" {
info["cluster address"] = coreConfig.ClusterAddr
infoKeys = append(infoKeys, "cluster address")
}
// Initialize the listeners
lns := make([]ServerListener, 0, len(config.Listeners))
for _, lnConfig := range config.Listeners {
ln, _, _, err := server.NewListener(lnConfig.Type, lnConfig.Config, c.logWriter, c.UI)
if err != nil {
c.UI.Error(fmt.Sprintf("Error initializing listener of type %s: %s", lnConfig.Type, err))
return 1
}
lns = append(lns, ServerListener{
Listener: ln,
config: lnConfig.Config,
})
}
listenerCloseFunc := func() {
for _, ln := range lns {
ln.Listener.Close()
}
}
defer c.cleanupGuard.Do(listenerCloseFunc)
infoKeys = append(infoKeys, "version")
verInfo := version.GetVersion()
info["version"] = verInfo.FullVersionNumber(false)
if verInfo.Revision != "" {
info["version sha"] = strings.Trim(verInfo.Revision, "'")
infoKeys = append(infoKeys, "version sha")
}
infoKeys = append(infoKeys, "recovery mode")
info["recovery mode"] = "true"
// Server configuration output
padding := 24
sort.Strings(infoKeys)
c.UI.Output("==> Vault server configuration:\n")
for _, k := range infoKeys {
c.UI.Output(fmt.Sprintf(
"%s%s: %s",
strings.Repeat(" ", padding-len(k)),
strings.Title(k),
info[k]))
}
c.UI.Output("")
for _, ln := range lns {
handler := vaulthttp.Handler(&vault.HandlerProperties{
Core: core,
MaxRequestSize: ln.maxRequestSize,
MaxRequestDuration: ln.maxRequestDuration,
DisablePrintableCheck: config.DisablePrintableCheck,
RecoveryMode: c.flagRecovery,
RecoveryToken: atomic.NewString(""),
})
server := &http.Server{
Handler: handler,
ReadHeaderTimeout: 10 * time.Second,
ReadTimeout: 30 * time.Second,
IdleTimeout: 5 * time.Minute,
ErrorLog: c.logger.StandardLogger(nil),
}
go server.Serve(ln.Listener)
}
if sealConfigError != nil {
init, err := core.Initialized(context.Background())
if err != nil {
c.UI.Error(fmt.Sprintf("Error checking if core is initialized: %v", err))
return 1
}
if init {
c.UI.Error("Vault is initialized but no Seal key could be loaded")
return 1
}
}
if newCoreError != nil {
c.UI.Warn(wrapAtLength(
"WARNING! A non-fatal error occurred during initialization. Please " +
"check the logs for more information."))
c.UI.Warn("")
}
if !c.flagCombineLogs {
c.UI.Output("==> Vault server started! Log data will stream in below:\n")
}
c.logGate.Flush()
for {
select {
case <-c.ShutdownCh:
c.UI.Output("==> Vault shutdown triggered")
c.cleanupGuard.Do(listenerCloseFunc)
if err := core.Shutdown(); err != nil {
c.UI.Error(fmt.Sprintf("Error with core shutdown: %s", err))
}
return 0
case <-c.SigUSR2Ch:
buf := make([]byte, 32*1024*1024)
n := runtime.Stack(buf[:], true)
c.logger.Info("goroutine trace", "stack", string(buf[:n]))
}
}
return 0
}
func (c *ServerCommand) adjustLogLevel(config *server.Config, logLevelWasNotSet bool) (string, error) {
var logLevelString string
if config.LogLevel != "" && logLevelWasNotSet {
configLogLevel := strings.ToLower(strings.TrimSpace(config.LogLevel))
logLevelString = configLogLevel
switch configLogLevel {
case "trace":
c.logger.SetLevel(log.Trace)
case "debug":
c.logger.SetLevel(log.Debug)
case "notice", "info", "":
c.logger.SetLevel(log.Info)
case "warn", "warning":
c.logger.SetLevel(log.Warn)
case "err", "error":
c.logger.SetLevel(log.Error)
default:
return "", fmt.Errorf("unknown log level: %s", config.LogLevel)
}
}
return logLevelString, nil
}
func (c *ServerCommand) processLogLevelAndFormat(config *server.Config) (log.Level, string, bool, logging.LogFormat, error) {
// Create a logger. We wrap it in a gated writer so that it doesn't
// start logging too early.
c.logGate = gatedwriter.NewWriter(os.Stderr)
c.logWriter = c.logGate
if c.flagCombineLogs {
c.logWriter = os.Stdout
}
var level log.Level
var logLevelWasNotSet bool
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
logFormat := logging.UnspecifiedFormat
logLevelString := c.flagLogLevel
c.flagLogLevel = strings.ToLower(strings.TrimSpace(c.flagLogLevel))
switch c.flagLogLevel {
case notSetValue, "":
logLevelWasNotSet = true
logLevelString = "info"
level = log.Info
case "trace":
level = log.Trace
case "debug":
level = log.Debug
case "notice", "info":
level = log.Info
case "warn", "warning":
level = log.Warn
case "err", "error":
level = log.Error
default:
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
return level, logLevelString, logLevelWasNotSet, logFormat, fmt.Errorf("unknown log level: %s", c.flagLogLevel)
}
if c.flagLogFormat != notSetValue {
var err error
logFormat, err = logging.ParseLogFormat(c.flagLogFormat)
if err != nil {
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
return level, logLevelString, logLevelWasNotSet, logFormat, err
}
}
if logFormat == logging.UnspecifiedFormat {
logFormat = logging.ParseEnvLogFormat()
}
if logFormat == logging.UnspecifiedFormat {
var err error
logFormat, err = logging.ParseLogFormat(config.LogFormat)
if err != nil {
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
return level, logLevelString, logLevelWasNotSet, logFormat, err
}
}
return level, logLevelString, logLevelWasNotSet, logFormat, nil
}
func (c *ServerCommand) Run(args []string) int {
f := c.Flags()
if err := f.Parse(args); err != nil {
c.UI.Error(err.Error())
return 1
}
if c.flagRecovery {
return c.runRecoveryMode()
}
// Automatically enable dev mode if other dev flags are provided.
if c.flagDevConsul || c.flagDevHA || c.flagDevTransactional || c.flagDevLeasedKV || c.flagDevThreeNode || c.flagDevFourCluster || c.flagDevAutoSeal || c.flagDevKVV1 {
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
c.flagDev = true
}
// Validation
if !c.flagDev {
switch {
case len(c.flagConfigs) == 0:
c.UI.Error("Must specify at least one config path using -config")
return 1
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
case c.flagDevRootTokenID != "":
c.UI.Warn(wrapAtLength(
"You cannot specify a custom root token ID outside of \"dev\" mode. " +
"Your request has been ignored."))
c.flagDevRootTokenID = ""
}
}
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
// Load the configuration
var config *server.Config
if c.flagDev {
var devStorageType string
switch {
case c.flagDevConsul:
devStorageType = "consul"
case c.flagDevHA && c.flagDevTransactional:
devStorageType = "inmem_transactional_ha"
case !c.flagDevHA && c.flagDevTransactional:
devStorageType = "inmem_transactional"
case c.flagDevHA && !c.flagDevTransactional:
devStorageType = "inmem_ha"
default:
devStorageType = "inmem"
}
config = server.DevConfig(devStorageType)
if c.flagDevListenAddr != "" {
config.Listeners[0].Config["address"] = c.flagDevListenAddr
}
}
parsedConfig, err := c.parseConfig()
if err != nil {
c.UI.Error(err.Error())
return 1
}
if config == nil {
config = parsedConfig
} else {
config = config.Merge(parsedConfig)
}
// Ensure at least one config was found.
if config == nil {
c.UI.Output(wrapAtLength(
"No configuration files found. Please provide configurations with the " +
"-config flag. If you are supplying the path to a directory, please " +
"ensure the directory contains files with the .hcl or .json " +
"extension."))
return 1
}
level, logLevelString, logLevelWasNotSet, logFormat, err := c.processLogLevelAndFormat(config)
if err != nil {
c.UI.Error(err.Error())
return 1
}
if c.flagDevThreeNode || c.flagDevFourCluster {
c.logger = log.New(&log.LoggerOptions{
Mutex: &sync.Mutex{},
Output: c.logWriter,
Level: log.Trace,
})
} else {
c.logger = log.New(&log.LoggerOptions{
Output: c.logWriter,
Level: level,
// Note that if logFormat is either unspecified or standard, then
// the resulting logger's format will be standard.
JSONFormat: logFormat == logging.JSONFormat,
})
}
allLoggers := []log.Logger{c.logger}
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
logLevelStr, err := c.adjustLogLevel(config, logLevelWasNotSet)
if err != nil {
c.UI.Error(err.Error())
return 1
}
if logLevelStr != "" {
logLevelString = logLevelStr
}
// create GRPC logger
namedGRPCLogFaker := c.logger.Named("grpclogfaker")
allLoggers = append(allLoggers, namedGRPCLogFaker)
grpclog.SetLogger(&grpclogFaker{
logger: namedGRPCLogFaker,
log: os.Getenv("VAULT_GRPC_LOGGING") != "",
})
if memProfilerEnabled {
c.startMemProfiler()
}
2015-05-21 00:49:16 +00:00
// Ensure that a backend is provided
if config.Storage == nil {
c.UI.Output("A storage backend must be specified")
return 1
}
if config.DefaultMaxRequestDuration != 0 {
vault.DefaultMaxRequestDuration = config.DefaultMaxRequestDuration
}
// log proxy settings
proxyCfg := httpproxy.FromEnvironment()
c.logger.Info("proxy environment", "http_proxy", proxyCfg.HTTPProxy,
"https_proxy", proxyCfg.HTTPSProxy, "no_proxy", proxyCfg.NoProxy)
// If mlockall(2) isn't supported, show a warning. We disable this in dev
// because it is quite scary to see when first using Vault. We also disable
// this if the user has explicitly disabled mlock in configuration.
if !c.flagDev && !config.DisableMlock && !mlock.Supported() {
c.UI.Warn(wrapAtLength(
"WARNING! mlock is not supported on this system! An mlockall(2)-like " +
"syscall to prevent memory from being swapped to disk is not " +
"supported on this system. For better security, only run Vault on " +
"systems where this call is supported. If you are running Vault " +
"in a Docker container, provide the IPC_LOCK cap to the container."))
2015-04-28 22:04:40 +00:00
}
metricsHelper, err := c.setupTelemetry(config)
if err != nil {
c.UI.Error(fmt.Sprintf("Error initializing telemetry: %s", err))
return 1
}
2015-03-13 17:09:38 +00:00
// Initialize the backend
factory, exists := c.PhysicalBackends[config.Storage.Type]
if !exists {
c.UI.Error(fmt.Sprintf("Unknown storage type %s", config.Storage.Type))
return 1
}
// Do any custom configuration needed per backend
switch config.Storage.Type {
case "consul":
if config.ServiceRegistration == nil {
// If Consul is configured for storage and service registration is unconfigured,
// use Consul for service registration without requiring additional configuration.
// This maintains backward-compatibility.
config.ServiceRegistration = &server.ServiceRegistration{
Type: "consul",
Config: config.Storage.Config,
}
}
case "raft":
if envCA := os.Getenv("VAULT_CLUSTER_ADDR"); envCA != "" {
config.ClusterAddr = envCA
}
if len(config.ClusterAddr) == 0 {
c.UI.Error("Cluster address must be set when using raft storage")
return 1
}
Raft Storage Backend (#6888) * Work on raft backend * Add logstore locally * Add encryptor and unsealable interfaces * Add clustering support to raft * Remove client and handler * Bootstrap raft on init * Cleanup raft logic a bit * More raft work * Work on TLS config * More work on bootstrapping * Fix build * More work on bootstrapping * More bootstrapping work * fix build * Remove consul dep * Fix build * merged oss/master into raft-storage * Work on bootstrapping * Get bootstrapping to work * Clean up FMS and node-id * Update local node ID logic * Cleanup node-id change * Work on snapshotting * Raft: Add remove peer API (#906) * Add remove peer API * Add some comments * Fix existing snapshotting (#909) * Raft get peers API (#912) * Read raft configuration * address review feedback * Use the Leadership Transfer API to step-down the active node (#918) * Raft join and unseal using Shamir keys (#917) * Raft join using shamir * Store AEAD instead of master key * Split the raft join process to answer the challenge after a successful unseal * get the follower to standby state * Make unseal work * minor changes * Some input checks * reuse the shamir seal access instead of new default seal access * refactor joinRaftSendAnswer function * Synchronously send answer in auto-unseal case * Address review feedback * Raft snapshots (#910) * Fix existing snapshotting * implement the noop snapshotting * Add comments and switch log libraries * add some snapshot tests * add snapshot test file * add TODO * More work on raft snapshotting * progress on the ConfigStore strategy * Don't use two buckets * Update the snapshot store logic to hide the file logic * Add more backend tests * Cleanup code a bit * [WIP] Raft recovery (#938) * Add recovery functionality * remove fmt.Printfs * Fix a few fsm bugs * Add max size value for raft backend (#942) * Add max size value for raft backend * Include physical.ErrValueTooLarge in the message * Raft snapshot Take/Restore API (#926) * Inital work on raft snapshot APIs * Always redirect snapshot install/download requests * More work on the snapshot APIs * Cleanup code a bit * On restore handle special cases * Use the seal to encrypt the sha sum file * Add sealer mechanism and fix some bugs * Call restore while state lock is held * Send restore cb trigger through raft log * Make error messages nicer * Add test helpers * Add snapshot test * Add shamir unseal test * Add more raft snapshot API tests * Fix locking * Change working to initalize * Add underlying raw object to test cluster core * Move leaderUUID to core * Add raft TLS rotation logic (#950) * Add TLS rotation logic * Cleanup logic a bit * Add/Remove from follower state on add/remove peer * add comments * Update more comments * Update request_forwarding_service.proto * Make sure we populate all nodes in the followerstate obj * Update times * Apply review feedback * Add more raft config setting (#947) * Add performance config setting * Add more config options and fix tests * Test Raft Recovery (#944) * Test raft recovery * Leave out a node during recovery * remove unused struct * Update physical/raft/snapshot_test.go * Update physical/raft/snapshot_test.go * fix vendoring * Switch to new raft interface * Remove unused files * Switch a gogo -> proto instance * Remove unneeded vault dep in go.sum * Update helper/testhelpers/testhelpers.go Co-Authored-By: Calvin Leung Huang <cleung2010@gmail.com> * Update vault/cluster/cluster.go * track active key within the keyring itself (#6915) * track active key within the keyring itself * lookup and store using the active key ID * update docstring * minor refactor * Small text fixes (#6912) * Update physical/raft/raft.go Co-Authored-By: Calvin Leung Huang <cleung2010@gmail.com> * review feedback * Move raft logical system into separate file * Update help text a bit * Enforce cluster addr is set and use it for raft bootstrapping * Fix tests * fix http test panic * Pull in latest raft-snapshot library * Add comment
2019-06-20 19:14:58 +00:00
}
namedStorageLogger := c.logger.Named("storage." + config.Storage.Type)
allLoggers = append(allLoggers, namedStorageLogger)
backend, err := factory(config.Storage.Config, namedStorageLogger)
2015-03-13 17:09:38 +00:00
if err != nil {
c.UI.Error(fmt.Sprintf("Error initializing storage of type %s: %s", config.Storage.Type, err))
2015-03-13 17:09:38 +00:00
return 1
}
// Prevent server startup if migration is active
2018-10-23 06:34:02 +00:00
if c.storageMigrationActive(backend) {
return 1
}
// Instantiate the wait group
c.WaitGroup = &sync.WaitGroup{}
// Initialize the Service Discovery, if there is one
var configSR sr.ServiceRegistration
if config.ServiceRegistration != nil {
sdFactory, ok := c.ServiceRegistrations[config.ServiceRegistration.Type]
if !ok {
c.UI.Error(fmt.Sprintf("Unknown service_registration type %s", config.ServiceRegistration.Type))
return 1
}
namedSDLogger := c.logger.Named("service_registration." + config.ServiceRegistration.Type)
allLoggers = append(allLoggers, namedSDLogger)
// Since we haven't even begun starting Vault's core yet,
// we know that Vault is in its pre-running state.
state := sr.State{
VaultVersion: version.GetVersion().VersionNumber(),
IsInitialized: false,
IsSealed: true,
IsActive: false,
IsPerformanceStandby: false,
}
configSR, err = sdFactory(config.ServiceRegistration.Config, namedSDLogger, state, config.Storage.RedirectAddr)
if err != nil {
c.UI.Error(fmt.Sprintf("Error initializing service_registration of type %s: %s", config.ServiceRegistration.Type, err))
return 1
}
if err := configSR.Run(c.ShutdownCh, c.WaitGroup); err != nil {
c.UI.Error(fmt.Sprintf("Error running service_registration of type %s: %s", config.ServiceRegistration.Type, err))
return 1
}
}
2016-04-04 17:46:33 +00:00
infoKeys := make([]string, 0, 10)
info := make(map[string]string)
info["log level"] = logLevelString
infoKeys = append(infoKeys, "log level")
2016-04-04 17:46:33 +00:00
var barrierSeal vault.Seal
var unwrapSeal vault.Seal
2018-09-18 03:03:00 +00:00
var sealConfigError error
if c.flagDevAutoSeal {
barrierSeal = vault.NewAutoSeal(vaultseal.NewTestSeal(nil))
} else {
// Handle the case where no seal is provided
switch len(config.Seals) {
case 0:
config.Seals = append(config.Seals, &server.Seal{Type: wrapping.Shamir})
case 1:
// If there's only one seal and it's disabled assume they want to
// migrate to a shamir seal and simply didn't provide it
if config.Seals[0].Disabled {
config.Seals = append(config.Seals, &server.Seal{Type: wrapping.Shamir})
}
}
for _, configSeal := range config.Seals {
sealType := wrapping.Shamir
if !configSeal.Disabled && os.Getenv("VAULT_SEAL_TYPE") != "" {
sealType = os.Getenv("VAULT_SEAL_TYPE")
configSeal.Type = sealType
} else {
sealType = configSeal.Type
}
var seal vault.Seal
sealLogger := c.logger.Named(sealType)
allLoggers = append(allLoggers, sealLogger)
seal, sealConfigError = serverseal.ConfigureSeal(configSeal, &infoKeys, &info, sealLogger, vault.NewDefaultSeal(&vaultseal.Access{
Wrapper: aeadwrapper.NewWrapper(&wrapping.WrapperOptions{
Logger: c.logger.Named("shamir"),
}),
}))
if sealConfigError != nil {
if !errwrap.ContainsType(sealConfigError, new(logical.KeyNotFoundError)) {
c.UI.Error(fmt.Sprintf(
"Error parsing Seal configuration: %s", sealConfigError))
return 1
}
}
if seal == nil {
c.UI.Error(fmt.Sprintf(
"After configuring seal nil returned, seal type was %s", sealType))
return 1
}
2016-04-04 17:46:33 +00:00
if configSeal.Disabled {
unwrapSeal = seal
} else {
barrierSeal = seal
2016-12-05 17:28:12 +00:00
}
// Ensure that the seal finalizer is called, even if using verify-only
defer func() {
err = seal.Finalize(context.Background())
if err != nil {
c.UI.Error(fmt.Sprintf("Error finalizing seals: %v", err))
}
}()
}
}
if barrierSeal == nil {
c.UI.Error(fmt.Sprintf("Could not create barrier seal! Most likely proper Seal configuration information was not set, but no error was generated."))
2016-12-05 17:28:12 +00:00
return 1
}
// prepare a secure random reader for core
secureRandomReader, err := createSecureRandomReaderFunc(config, &barrierSeal)
if err != nil {
c.UI.Error(err.Error())
return 1
}
2015-12-14 22:58:30 +00:00
coreConfig := &vault.CoreConfig{
RawConfig: config,
2018-09-18 03:03:00 +00:00
Physical: backend,
RedirectAddr: config.Storage.RedirectAddr,
StorageType: config.Storage.Type,
2018-09-18 03:03:00 +00:00
HAPhysical: nil,
ServiceRegistration: configSR,
Seal: barrierSeal,
2018-09-18 03:03:00 +00:00
AuditBackends: c.AuditBackends,
CredentialBackends: c.CredentialBackends,
LogicalBackends: c.LogicalBackends,
Logger: c.logger,
DisableCache: config.DisableCache,
DisableMlock: config.DisableMlock,
MaxLeaseTTL: config.MaxLeaseTTL,
DefaultLeaseTTL: config.DefaultLeaseTTL,
ClusterName: config.ClusterName,
CacheSize: config.CacheSize,
PluginDirectory: config.PluginDirectory,
EnableUI: config.EnableUI,
EnableRaw: config.EnableRawEndpoint,
DisableSealWrap: config.DisableSealWrap,
DisablePerformanceStandby: config.DisablePerformanceStandby,
2018-10-23 19:03:17 +00:00
DisableIndexing: config.DisableIndexing,
2018-09-18 03:03:00 +00:00
AllLoggers: allLoggers,
2018-11-07 01:21:24 +00:00
BuiltinRegistry: builtinplugins.Registry,
DisableKeyEncodingChecks: config.DisablePrintableCheck,
MetricsHelper: metricsHelper,
SecureRandomReader: secureRandomReader,
2015-12-14 22:58:30 +00:00
}
if c.flagDev {
coreConfig.DevToken = c.flagDevRootTokenID
if c.flagDevLeasedKV {
coreConfig.LogicalBackends["kv"] = vault.LeasedPassthroughBackendFactory
}
if c.flagDevPluginDir != "" {
coreConfig.PluginDirectory = c.flagDevPluginDir
}
if c.flagDevLatency > 0 {
injectLatency := time.Duration(c.flagDevLatency) * time.Millisecond
if _, txnOK := backend.(physical.Transactional); txnOK {
coreConfig.Physical = physical.NewTransactionalLatencyInjector(backend, injectLatency, c.flagDevLatencyJitter, c.logger)
} else {
coreConfig.Physical = physical.NewLatencyInjector(backend, injectLatency, c.flagDevLatencyJitter, c.logger)
}
}
}
if c.flagDevThreeNode {
return c.enableThreeNodeDevCluster(coreConfig, info, infoKeys, c.flagDevListenAddr, os.Getenv("VAULT_DEV_TEMP_DIR"))
}
2018-09-18 03:03:00 +00:00
if c.flagDevFourCluster {
return enableFourClusterDev(c, coreConfig, info, infoKeys, c.flagDevListenAddr, os.Getenv("VAULT_DEV_TEMP_DIR"))
2018-09-18 03:03:00 +00:00
}
var disableClustering bool
// Initialize the separate HA storage backend, if it exists
var ok bool
if config.HAStorage != nil {
factory, exists := c.PhysicalBackends[config.HAStorage.Type]
if !exists {
c.UI.Error(fmt.Sprintf("Unknown HA storage type %s", config.HAStorage.Type))
return 1
}
habackend, err := factory(config.HAStorage.Config, c.logger)
if err != nil {
c.UI.Error(fmt.Sprintf(
"Error initializing HA storage of type %s: %s", config.HAStorage.Type, err))
return 1
}
2015-12-14 22:58:30 +00:00
if coreConfig.HAPhysical, ok = habackend.(physical.HABackend); !ok {
c.UI.Error("Specified HA storage does not support HA")
return 1
}
if !coreConfig.HAPhysical.HAEnabled() {
c.UI.Error("Specified HA storage has HA support disabled; please consult documentation")
return 1
}
coreConfig.RedirectAddr = config.HAStorage.RedirectAddr
disableClustering = config.HAStorage.DisableClustering
if !disableClustering {
coreConfig.ClusterAddr = config.HAStorage.ClusterAddr
}
} else {
if coreConfig.HAPhysical, ok = backend.(physical.HABackend); ok {
coreConfig.RedirectAddr = config.Storage.RedirectAddr
disableClustering = config.Storage.DisableClustering
if !disableClustering {
coreConfig.ClusterAddr = config.Storage.ClusterAddr
}
}
}
if envRA := os.Getenv("VAULT_API_ADDR"); envRA != "" {
coreConfig.RedirectAddr = envRA
} else if envRA := os.Getenv("VAULT_REDIRECT_ADDR"); envRA != "" {
coreConfig.RedirectAddr = envRA
} else if envAA := os.Getenv("VAULT_ADVERTISE_ADDR"); envAA != "" {
coreConfig.RedirectAddr = envAA
}
// Attempt to detect the redirect address, if possible
if coreConfig.RedirectAddr == "" {
c.logger.Warn("no `api_addr` value specified in config or in VAULT_API_ADDR; falling back to detection if possible, but this value should be manually set")
}
var detect physical.RedirectDetect
if coreConfig.HAPhysical != nil && coreConfig.HAPhysical.HAEnabled() {
detect, ok = coreConfig.HAPhysical.(physical.RedirectDetect)
} else {
detect, ok = coreConfig.Physical.(physical.RedirectDetect)
}
if ok && coreConfig.RedirectAddr == "" {
redirect, err := c.detectRedirect(detect, config)
if err != nil {
c.UI.Error(fmt.Sprintf("Error detecting api address: %s", err))
} else if redirect == "" {
c.UI.Error("Failed to detect api address")
} else {
coreConfig.RedirectAddr = redirect
}
}
if coreConfig.RedirectAddr == "" && c.flagDev {
2017-02-24 15:45:29 +00:00
coreConfig.RedirectAddr = fmt.Sprintf("http://%s", config.Listeners[0].Config["address"])
}
// After the redirect bits are sorted out, if no cluster address was
// explicitly given, derive one from the redirect addr
if disableClustering {
coreConfig.ClusterAddr = ""
} else if envCA := os.Getenv("VAULT_CLUSTER_ADDR"); envCA != "" {
coreConfig.ClusterAddr = envCA
2017-02-24 15:45:29 +00:00
} else {
var addrToUse string
switch {
case coreConfig.ClusterAddr == "" && coreConfig.RedirectAddr != "":
2017-02-24 15:45:29 +00:00
addrToUse = coreConfig.RedirectAddr
case c.flagDev:
2017-02-24 15:45:29 +00:00
addrToUse = fmt.Sprintf("http://%s", config.Listeners[0].Config["address"])
default:
goto CLUSTER_SYNTHESIS_COMPLETE
2017-02-24 15:45:29 +00:00
}
u, err := url.ParseRequestURI(addrToUse)
if err != nil {
c.UI.Error(fmt.Sprintf(
"Error parsing synthesized cluster address %s: %v", addrToUse, err))
2016-04-04 14:44:22 +00:00
return 1
}
host, port, err := net.SplitHostPort(u.Host)
2015-03-31 23:44:47 +00:00
if err != nil {
// This sucks, as it's a const in the function but not exported in the package
if strings.Contains(err.Error(), "missing port in address") {
host = u.Host
port = "443"
} else {
c.UI.Error(fmt.Sprintf("Error parsing api address: %v", err))
return 1
}
}
nPort, err := strconv.Atoi(port)
if err != nil {
c.UI.Error(fmt.Sprintf(
"Error parsing synthesized address; failed to convert %q to a numeric: %v", port, err))
2015-03-31 23:44:47 +00:00
return 1
}
u.Host = net.JoinHostPort(host, strconv.Itoa(nPort+1))
// Will always be TLS-secured
u.Scheme = "https"
coreConfig.ClusterAddr = u.String()
}
CLUSTER_SYNTHESIS_COMPLETE:
if coreConfig.RedirectAddr == coreConfig.ClusterAddr && len(coreConfig.RedirectAddr) != 0 {
c.UI.Error(fmt.Sprintf(
"Address %q used for both API and cluster addresses", coreConfig.RedirectAddr))
return 1
}
if coreConfig.ClusterAddr != "" {
// Force https as we'll always be TLS-secured
u, err := url.ParseRequestURI(coreConfig.ClusterAddr)
if err != nil {
2018-04-17 12:36:38 +00:00
c.UI.Error(fmt.Sprintf("Error parsing cluster address %s: %v", coreConfig.ClusterAddr, err))
return 11
}
u.Scheme = "https"
coreConfig.ClusterAddr = u.String()
}
// Override the UI enabling config by the environment variable
if enableUI := os.Getenv("VAULT_UI"); enableUI != "" {
var err error
coreConfig.EnableUI, err = strconv.ParseBool(enableUI)
if err != nil {
c.UI.Output("Error parsing the environment variable VAULT_UI")
return 1
}
}
// If ServiceRegistration is configured, then the backend must support HA
isBackendHA := coreConfig.HAPhysical != nil && coreConfig.HAPhysical.HAEnabled()
if (coreConfig.ServiceRegistration != nil) && !isBackendHA {
c.UI.Output("service_registration is configured, but storage does not support HA")
return 1
}
// Initialize the core
core, newCoreError := vault.NewCore(coreConfig)
if newCoreError != nil {
if vault.IsFatalError(newCoreError) {
c.UI.Error(fmt.Sprintf("Error initializing core: %s", newCoreError))
return 1
}
2015-03-31 23:44:47 +00:00
}
2016-09-30 04:06:40 +00:00
// Copy the reload funcs pointers back
c.reloadFuncs = coreConfig.ReloadFuncs
c.reloadFuncsLock = coreConfig.ReloadFuncsLock
2015-04-04 19:06:41 +00:00
// Compile server information for output later
info["storage"] = config.Storage.Type
info["mlock"] = fmt.Sprintf(
"supported: %v, enabled: %v",
mlock.Supported(), !config.DisableMlock && mlock.Supported())
infoKeys = append(infoKeys, "mlock", "storage")
2015-04-04 19:06:41 +00:00
2017-02-24 15:45:29 +00:00
if coreConfig.ClusterAddr != "" {
info["cluster address"] = coreConfig.ClusterAddr
infoKeys = append(infoKeys, "cluster address")
}
if coreConfig.RedirectAddr != "" {
info["api address"] = coreConfig.RedirectAddr
infoKeys = append(infoKeys, "api address")
2017-02-24 15:45:29 +00:00
}
if config.HAStorage != nil {
info["HA storage"] = config.HAStorage.Type
infoKeys = append(infoKeys, "HA storage")
} else {
// If the storage supports HA, then note it
2015-12-14 22:58:30 +00:00
if coreConfig.HAPhysical != nil {
if coreConfig.HAPhysical.HAEnabled() {
info["storage"] += " (HA available)"
} else {
info["storage"] += " (HA disabled)"
}
}
}
2016-08-19 15:03:53 +00:00
clusterAddrs := []*net.TCPAddr{}
2015-03-12 22:30:07 +00:00
// Initialize the listeners
2018-04-17 22:52:09 +00:00
lns := make([]ServerListener, 0, len(config.Listeners))
2016-09-30 04:06:40 +00:00
c.reloadFuncsLock.Lock()
2015-04-04 19:06:41 +00:00
for i, lnConfig := range config.Listeners {
ln, props, reloadFunc, err := server.NewListener(lnConfig.Type, lnConfig.Config, c.logWriter, c.UI)
2015-03-13 17:09:38 +00:00
if err != nil {
c.UI.Error(fmt.Sprintf("Error initializing listener of type %s: %s", lnConfig.Type, err))
2015-03-13 17:09:38 +00:00
return 1
}
if reloadFunc != nil {
2016-09-30 04:06:40 +00:00
relSlice := (*c.reloadFuncs)["listener|"+lnConfig.Type]
relSlice = append(relSlice, reloadFunc)
2016-09-30 04:06:40 +00:00
(*c.reloadFuncs)["listener|"+lnConfig.Type] = relSlice
}
if !disableClustering && lnConfig.Type == "tcp" {
var addrRaw interface{}
var addr string
var ok bool
if addrRaw, ok = lnConfig.Config["cluster_address"]; ok {
addr = addrRaw.(string)
2016-08-19 15:03:53 +00:00
tcpAddr, err := net.ResolveTCPAddr("tcp", addr)
if err != nil {
c.UI.Error(fmt.Sprintf("Error resolving cluster_address: %s", err))
2016-08-19 15:03:53 +00:00
return 1
}
clusterAddrs = append(clusterAddrs, tcpAddr)
} else {
tcpAddr, ok := ln.Addr().(*net.TCPAddr)
if !ok {
c.UI.Error("Failed to parse tcp listener")
return 1
}
2017-02-24 15:45:29 +00:00
clusterAddr := &net.TCPAddr{
2016-08-19 15:03:53 +00:00
IP: tcpAddr.IP,
Port: tcpAddr.Port + 1,
2017-02-24 15:45:29 +00:00
}
clusterAddrs = append(clusterAddrs, clusterAddr)
addr = clusterAddr.String()
}
props["cluster address"] = addr
}
var maxRequestSize int64 = vaulthttp.DefaultMaxRequestSize
if valRaw, ok := lnConfig.Config["max_request_size"]; ok {
val, err := parseutil.ParseInt(valRaw)
if err != nil {
c.UI.Error(fmt.Sprintf("Could not parse max_request_size value %v", valRaw))
return 1
}
if val >= 0 {
maxRequestSize = val
}
}
props["max_request_size"] = fmt.Sprintf("%d", maxRequestSize)
maxRequestDuration := vault.DefaultMaxRequestDuration
if valRaw, ok := lnConfig.Config["max_request_duration"]; ok {
val, err := parseutil.ParseDurationSecond(valRaw)
if err != nil {
c.UI.Error(fmt.Sprintf("Could not parse max_request_duration value %v", valRaw))
return 1
}
if val >= 0 {
maxRequestDuration = val
}
}
props["max_request_duration"] = fmt.Sprintf("%s", maxRequestDuration.String())
var unauthenticatedMetricsAccess bool
if telemetryRaw, ok := lnConfig.Config["telemetry"]; ok {
telemetry, ok := telemetryRaw.([]map[string]interface{})
if !ok {
c.UI.Error(fmt.Sprintf("Could not parse telemetry sink value %v", telemetryRaw))
return 1
}
for _, item := range telemetry {
if valRaw, ok := item["unauthenticated_metrics_access"]; ok {
unauthenticatedMetricsAccess, err = parseutil.ParseBool(valRaw)
if err != nil {
c.UI.Error(fmt.Sprintf("Could not parse unauthenticated_metrics_access value %v", valRaw))
return 1
}
}
}
}
lns = append(lns, ServerListener{
Listener: ln,
config: lnConfig.Config,
maxRequestSize: maxRequestSize,
maxRequestDuration: maxRequestDuration,
unauthenticatedMetricsAccess: unauthenticatedMetricsAccess,
})
2015-04-04 19:06:41 +00:00
// Store the listener props for output later
key := fmt.Sprintf("listener %d", i+1)
propsList := make([]string, 0, len(props))
for k, v := range props {
propsList = append(propsList, fmt.Sprintf(
"%s: %q", k, v))
}
sort.Strings(propsList)
infoKeys = append(infoKeys, key)
info[key] = fmt.Sprintf(
"%s (%s)", lnConfig.Type, strings.Join(propsList, ", "))
}
2016-09-30 04:06:40 +00:00
c.reloadFuncsLock.Unlock()
if !disableClustering {
if c.logger.IsDebug() {
c.logger.Debug("cluster listener addresses synthesized", "cluster_addresses", clusterAddrs)
2016-08-19 20:45:17 +00:00
}
2015-03-13 17:09:38 +00:00
}
2016-06-02 16:40:25 +00:00
// Make sure we close all listeners from this point on
listenerCloseFunc := func() {
2016-06-02 16:40:25 +00:00
for _, ln := range lns {
2018-04-17 22:52:09 +00:00
ln.Listener.Close()
2016-06-02 16:40:25 +00:00
}
}
defer c.cleanupGuard.Do(listenerCloseFunc)
2016-06-02 16:40:25 +00:00
2016-11-28 00:16:44 +00:00
infoKeys = append(infoKeys, "version")
verInfo := version.GetVersion()
info["version"] = verInfo.FullVersionNumber(false)
2016-11-28 00:16:44 +00:00
if verInfo.Revision != "" {
info["version sha"] = strings.Trim(verInfo.Revision, "'")
infoKeys = append(infoKeys, "version sha")
}
infoKeys = append(infoKeys, "cgo")
info["cgo"] = "disabled"
if version.CgoEnabled {
info["cgo"] = "enabled"
2016-11-28 00:16:44 +00:00
}
2015-11-09 18:52:55 +00:00
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
infoKeys = append(infoKeys, "recovery mode")
info["recovery mode"] = "false"
2015-04-04 19:06:41 +00:00
// Server configuration output
padding := 24
sort.Strings(infoKeys)
c.UI.Output("==> Vault server configuration:\n")
2015-04-04 19:06:41 +00:00
for _, k := range infoKeys {
c.UI.Output(fmt.Sprintf(
2015-04-04 19:06:41 +00:00
"%s%s: %s",
strings.Repeat(" ", padding-len(k)),
strings.Title(k),
info[k]))
}
c.UI.Output("")
2015-04-04 19:06:41 +00:00
// Tests might not want to start a vault server and just want to verify
// the configuration.
if c.flagTestVerifyOnly {
return 0
}
2017-10-10 16:27:51 +00:00
// This needs to happen before we first unseal, so before we trigger dev
// mode if it's set
core.SetClusterListenerAddrs(clusterAddrs)
core.SetClusterHandler(vaulthttp.Handler(&vault.HandlerProperties{
Core: core,
}))
2017-10-10 16:27:51 +00:00
2018-10-23 06:34:02 +00:00
// Before unsealing with stored keys, setup seal migration if needed
Raft Storage Backend (#6888) * Work on raft backend * Add logstore locally * Add encryptor and unsealable interfaces * Add clustering support to raft * Remove client and handler * Bootstrap raft on init * Cleanup raft logic a bit * More raft work * Work on TLS config * More work on bootstrapping * Fix build * More work on bootstrapping * More bootstrapping work * fix build * Remove consul dep * Fix build * merged oss/master into raft-storage * Work on bootstrapping * Get bootstrapping to work * Clean up FMS and node-id * Update local node ID logic * Cleanup node-id change * Work on snapshotting * Raft: Add remove peer API (#906) * Add remove peer API * Add some comments * Fix existing snapshotting (#909) * Raft get peers API (#912) * Read raft configuration * address review feedback * Use the Leadership Transfer API to step-down the active node (#918) * Raft join and unseal using Shamir keys (#917) * Raft join using shamir * Store AEAD instead of master key * Split the raft join process to answer the challenge after a successful unseal * get the follower to standby state * Make unseal work * minor changes * Some input checks * reuse the shamir seal access instead of new default seal access * refactor joinRaftSendAnswer function * Synchronously send answer in auto-unseal case * Address review feedback * Raft snapshots (#910) * Fix existing snapshotting * implement the noop snapshotting * Add comments and switch log libraries * add some snapshot tests * add snapshot test file * add TODO * More work on raft snapshotting * progress on the ConfigStore strategy * Don't use two buckets * Update the snapshot store logic to hide the file logic * Add more backend tests * Cleanup code a bit * [WIP] Raft recovery (#938) * Add recovery functionality * remove fmt.Printfs * Fix a few fsm bugs * Add max size value for raft backend (#942) * Add max size value for raft backend * Include physical.ErrValueTooLarge in the message * Raft snapshot Take/Restore API (#926) * Inital work on raft snapshot APIs * Always redirect snapshot install/download requests * More work on the snapshot APIs * Cleanup code a bit * On restore handle special cases * Use the seal to encrypt the sha sum file * Add sealer mechanism and fix some bugs * Call restore while state lock is held * Send restore cb trigger through raft log * Make error messages nicer * Add test helpers * Add snapshot test * Add shamir unseal test * Add more raft snapshot API tests * Fix locking * Change working to initalize * Add underlying raw object to test cluster core * Move leaderUUID to core * Add raft TLS rotation logic (#950) * Add TLS rotation logic * Cleanup logic a bit * Add/Remove from follower state on add/remove peer * add comments * Update more comments * Update request_forwarding_service.proto * Make sure we populate all nodes in the followerstate obj * Update times * Apply review feedback * Add more raft config setting (#947) * Add performance config setting * Add more config options and fix tests * Test Raft Recovery (#944) * Test raft recovery * Leave out a node during recovery * remove unused struct * Update physical/raft/snapshot_test.go * Update physical/raft/snapshot_test.go * fix vendoring * Switch to new raft interface * Remove unused files * Switch a gogo -> proto instance * Remove unneeded vault dep in go.sum * Update helper/testhelpers/testhelpers.go Co-Authored-By: Calvin Leung Huang <cleung2010@gmail.com> * Update vault/cluster/cluster.go * track active key within the keyring itself (#6915) * track active key within the keyring itself * lookup and store using the active key ID * update docstring * minor refactor * Small text fixes (#6912) * Update physical/raft/raft.go Co-Authored-By: Calvin Leung Huang <cleung2010@gmail.com> * review feedback * Move raft logical system into separate file * Update help text a bit * Enforce cluster addr is set and use it for raft bootstrapping * Fix tests * fix http test panic * Pull in latest raft-snapshot library * Add comment
2019-06-20 19:14:58 +00:00
if err := adjustCoreForSealMigration(c.logger, core, barrierSeal, unwrapSeal); err != nil {
2018-10-23 06:34:02 +00:00
c.UI.Error(err.Error())
return 1
}
// Attempt unsealing in a background goroutine. This is needed for when a
// Vault cluster with multiple servers is configured with auto-unseal but is
// uninitialized. Once one server initializes the storage backend, this
// goroutine will pick up the unseal keys and unseal this instance.
if !core.IsInSealMigration() {
go func() {
for {
err := core.UnsealWithStoredKeys(context.Background())
if err == nil {
return
}
if vault.IsFatalError(err) {
c.logger.Error("error unsealing core", "error", err)
return
} else {
c.logger.Warn("failed to unseal core", "error", err)
}
select {
case <-c.ShutdownCh:
return
case <-time.After(5 * time.Second):
}
}
}()
}
2017-10-23 21:39:21 +00:00
// When the underlying storage is raft, kick off retry join if it was specified
// in the configuration
if config.Storage.Type == "raft" {
if err := core.InitiateRetryJoin(context.Background()); err != nil {
c.UI.Error(fmt.Sprintf("Failed to initiate raft retry join, %q", err.Error()))
return 1
}
}
// Perform initialization of HTTP server after the verifyOnly check.
// If we're in Dev mode, then initialize the core
if c.flagDev && !c.flagDevSkipInit {
init, err := c.enableDev(core, coreConfig)
if err != nil {
c.UI.Error(fmt.Sprintf("Error initializing Dev mode: %s", err))
return 1
}
var plugins, pluginsNotLoaded []string
if c.flagDevPluginDir != "" && c.flagDevPluginInit {
2018-11-07 01:21:24 +00:00
f, err := os.Open(c.flagDevPluginDir)
if err != nil {
c.UI.Error(fmt.Sprintf("Error reading plugin dir: %s", err))
return 1
}
list, err := f.Readdirnames(0)
f.Close()
if err != nil {
c.UI.Error(fmt.Sprintf("Error listing plugins: %s", err))
return 1
}
for _, name := range list {
path := filepath.Join(f.Name(), name)
if err := c.addPlugin(path, init.RootToken, core); err != nil {
if !errwrap.Contains(err, vault.ErrPluginBadType.Error()) {
c.UI.Error(fmt.Sprintf("Error enabling plugin %s: %s", name, err))
return 1
}
pluginsNotLoaded = append(pluginsNotLoaded, name)
continue
}
plugins = append(plugins, name)
}
sort.Strings(plugins)
}
// Print the big dev mode warning!
c.UI.Warn(wrapAtLength(
"WARNING! dev mode is enabled! In this mode, Vault runs entirely " +
"in-memory and starts unsealed with a single unseal key. The root " +
"token is already authenticated to the CLI, so you can immediately " +
"begin using Vault."))
c.UI.Warn("")
c.UI.Warn("You may need to set the following environment variable:")
c.UI.Warn("")
endpointURL := "http://" + config.Listeners[0].Config["address"].(string)
if runtime.GOOS == "windows" {
c.UI.Warn("PowerShell:")
c.UI.Warn(fmt.Sprintf(" $env:VAULT_ADDR=\"%s\"", endpointURL))
c.UI.Warn("cmd.exe:")
c.UI.Warn(fmt.Sprintf(" set VAULT_ADDR=%s", endpointURL))
} else {
c.UI.Warn(fmt.Sprintf(" $ export VAULT_ADDR='%s'", endpointURL))
}
// Unseal key is not returned if stored shares is supported
if len(init.SecretShares) > 0 {
c.UI.Warn("")
c.UI.Warn(wrapAtLength(
"The unseal key and root token are displayed below in case you want " +
"to seal/unseal the Vault or re-authenticate."))
c.UI.Warn("")
c.UI.Warn(fmt.Sprintf("Unseal Key: %s", base64.StdEncoding.EncodeToString(init.SecretShares[0])))
}
if len(init.RecoveryShares) > 0 {
c.UI.Warn("")
c.UI.Warn(wrapAtLength(
"The recovery key and root token are displayed below in case you want " +
"to seal/unseal the Vault or re-authenticate."))
c.UI.Warn("")
c.UI.Warn(fmt.Sprintf("Recovery Key: %s", base64.StdEncoding.EncodeToString(init.RecoveryShares[0])))
}
c.UI.Warn(fmt.Sprintf("Root Token: %s", init.RootToken))
if len(plugins) > 0 {
c.UI.Warn("")
c.UI.Warn(wrapAtLength(
"The following dev plugins are registered in the catalog:"))
for _, p := range plugins {
c.UI.Warn(fmt.Sprintf(" - %s", p))
}
}
if len(pluginsNotLoaded) > 0 {
c.UI.Warn("")
c.UI.Warn(wrapAtLength(
"The following dev plugins FAILED to be registered in the catalog due to unknown type:"))
for _, p := range pluginsNotLoaded {
c.UI.Warn(fmt.Sprintf(" - %s", p))
}
}
c.UI.Warn("")
c.UI.Warn(wrapAtLength(
"Development mode should NOT be used in production installations!"))
c.UI.Warn("")
}
// Initialize the HTTP servers
for _, ln := range lns {
handler := vaulthttp.Handler(&vault.HandlerProperties{
Core: core,
MaxRequestSize: ln.maxRequestSize,
MaxRequestDuration: ln.maxRequestDuration,
DisablePrintableCheck: config.DisablePrintableCheck,
UnauthenticatedMetricsAccess: ln.unauthenticatedMetricsAccess,
Recovery Mode (#7559) * Initial work * rework * s/dr/recovery * Add sys/raw support to recovery mode (#7577) * Factor the raw paths out so they can be run with a SystemBackend. # Conflicts: # vault/logical_system.go * Add handleLogicalRecovery which is like handleLogical but is only sufficient for use with the sys-raw endpoint in recovery mode. No authentication is done yet. * Integrate with recovery-mode. We now handle unauthenticated sys/raw requests, albeit on path v1/raw instead v1/sys/raw. * Use sys/raw instead raw during recovery. * Don't bother persisting the recovery token. Authenticate sys/raw requests with it. * RecoveryMode: Support generate-root for autounseals (#7591) * Recovery: Abstract config creation and log settings * Recovery mode integration test. (#7600) * Recovery: Touch up (#7607) * Recovery: Touch up * revert the raw backend creation changes * Added recovery operation token prefix * Move RawBackend to its own file * Update API path and hit it using CLI flag on generate-root * Fix a panic triggered when handling a request that yields a nil response. (#7618) * Improve integ test to actually make changes while in recovery mode and verify they're still there after coming back in regular mode. * Refuse to allow a second recovery token to be generated. * Resize raft cluster to size 1 and start as leader (#7626) * RecoveryMode: Setup raft cluster post unseal (#7635) * Setup raft cluster post unseal in recovery mode * Remove marking as unsealed as its not needed * Address review comments * Accept only one seal config in recovery mode as there is no scope for migration
2019-10-15 04:55:31 +00:00
RecoveryMode: c.flagRecovery,
})
2018-04-17 22:52:09 +00:00
// We perform validation on the config earlier, we can just cast here
if _, ok := ln.config["x_forwarded_for_authorized_addrs"]; ok {
hopSkips := ln.config["x_forwarded_for_hop_skips"].(int)
authzdAddrs := ln.config["x_forwarded_for_authorized_addrs"].([]*sockaddr.SockAddrMarshaler)
rejectNotPresent := ln.config["x_forwarded_for_reject_not_present"].(bool)
rejectNonAuthz := ln.config["x_forwarded_for_reject_not_authorized"].(bool)
if len(authzdAddrs) > 0 {
handler = vaulthttp.WrapForwardedForHandler(handler, authzdAddrs, rejectNotPresent, rejectNonAuthz, hopSkips)
}
}
// server defaults
server := &http.Server{
Handler: handler,
ReadHeaderTimeout: 10 * time.Second,
ReadTimeout: 30 * time.Second,
IdleTimeout: 5 * time.Minute,
ErrorLog: c.logger.StandardLogger(nil),
}
// override server defaults with config values for read/write/idle timeouts if configured
if readHeaderTimeoutInterface, ok := ln.config["http_read_header_timeout"]; ok {
readHeaderTimeout, err := parseutil.ParseDurationSecond(readHeaderTimeoutInterface)
if err != nil {
c.UI.Error(fmt.Sprintf("Could not parse a time value for http_read_header_timeout %v", readHeaderTimeout))
return 1
}
server.ReadHeaderTimeout = readHeaderTimeout
}
if readTimeoutInterface, ok := ln.config["http_read_timeout"]; ok {
readTimeout, err := parseutil.ParseDurationSecond(readTimeoutInterface)
if err != nil {
c.UI.Error(fmt.Sprintf("Could not parse a time value for http_read_timeout %v", readTimeout))
return 1
}
server.ReadTimeout = readTimeout
}
if writeTimeoutInterface, ok := ln.config["http_write_timeout"]; ok {
writeTimeout, err := parseutil.ParseDurationSecond(writeTimeoutInterface)
if err != nil {
c.UI.Error(fmt.Sprintf("Could not parse a time value for http_write_timeout %v", writeTimeout))
return 1
}
server.WriteTimeout = writeTimeout
}
if idleTimeoutInterface, ok := ln.config["http_idle_timeout"]; ok {
idleTimeout, err := parseutil.ParseDurationSecond(idleTimeoutInterface)
if err != nil {
c.UI.Error(fmt.Sprintf("Could not parse a time value for http_idle_timeout %v", idleTimeout))
return 1
}
server.IdleTimeout = idleTimeout
}
// server config tests can exit now
if c.flagTestServerConfig {
continue
}
2018-04-17 22:52:09 +00:00
go server.Serve(ln.Listener)
}
if c.flagTestServerConfig {
return 0
}
2018-09-18 03:03:00 +00:00
if sealConfigError != nil {
init, err := core.Initialized(context.Background())
if err != nil {
c.UI.Error(fmt.Sprintf("Error checking if core is initialized: %v", err))
return 1
}
if init {
c.UI.Error("Vault is initialized but no Seal key could be loaded")
return 1
}
}
2016-04-04 14:44:22 +00:00
if newCoreError != nil {
c.UI.Warn(wrapAtLength(
"WARNING! A non-fatal error occurred during initialization. Please " +
"check the logs for more information."))
c.UI.Warn("")
2016-04-04 14:44:22 +00:00
}
2015-04-04 19:06:41 +00:00
// Output the header that the server has started
if !c.flagCombineLogs {
c.UI.Output("==> Vault server started! Log data will stream in below:\n")
}
// Inform any tests that the server is ready
select {
case c.startedCh <- struct{}{}:
default:
}
2015-04-04 19:06:41 +00:00
// Release the log gate.
c.logGate.Flush()
2015-04-04 19:06:41 +00:00
// Write out the PID to the file now that server has successfully started
if err := c.storePidFile(config.PidFile); err != nil {
c.UI.Error(fmt.Sprintf("Error storing PID: %s", err))
return 1
}
defer func() {
if err := c.removePidFile(config.PidFile); err != nil {
c.UI.Error(fmt.Sprintf("Error deleting the PID file: %s", err))
}
}()
// Wait for shutdown
shutdownTriggered := false
2016-03-11 22:01:26 +00:00
for !shutdownTriggered {
select {
case <-c.ShutdownCh:
c.UI.Output("==> Vault shutdown triggered")
2018-03-20 18:54:10 +00:00
// Stop the listeners so that we don't process further client requests.
c.cleanupGuard.Do(listenerCloseFunc)
// Shutdown will wait until after Vault is sealed, which means the
// request forwarding listeners will also be closed (and also
// waited for).
if err := core.Shutdown(); err != nil {
c.UI.Error(fmt.Sprintf("Error with core shutdown: %s", err))
}
shutdownTriggered = true
case <-c.SighupCh:
c.UI.Output("==> Vault reload triggered")
// Check for new log level
var config *server.Config
var level log.Level
for _, path := range c.flagConfigs {
current, err := server.LoadConfig(path)
if err != nil {
c.logger.Error("could not reload config", "path", path, "error", err)
goto RUNRELOADFUNCS
}
if config == nil {
config = current
} else {
config = config.Merge(current)
}
}
// Ensure at least one config was found.
if config == nil {
c.logger.Error("no config found at reload time")
goto RUNRELOADFUNCS
}
core.SetConfig(config)
if config.LogLevel != "" {
configLogLevel := strings.ToLower(strings.TrimSpace(config.LogLevel))
switch configLogLevel {
case "trace":
level = log.Trace
case "debug":
level = log.Debug
case "notice", "info", "":
level = log.Info
case "warn", "warning":
level = log.Warn
case "err", "error":
level = log.Error
default:
c.logger.Error("unknown log level found on reload", "level", config.LogLevel)
goto RUNRELOADFUNCS
}
core.SetLogLevel(level)
}
RUNRELOADFUNCS:
if err := c.Reload(c.reloadFuncsLock, c.reloadFuncs, c.flagConfigs); err != nil {
c.UI.Error(fmt.Sprintf("Error(s) were encountered during reload: %s", err))
}
case <-c.SigUSR2Ch:
buf := make([]byte, 32*1024*1024)
n := runtime.Stack(buf[:], true)
c.logger.Info("goroutine trace", "stack", string(buf[:n]))
}
}
2016-08-01 15:15:25 +00:00
// Wait for dependent goroutines to complete
c.WaitGroup.Wait()
2015-03-12 22:21:11 +00:00
return 0
}
func (c *ServerCommand) enableDev(core *vault.Core, coreConfig *vault.CoreConfig) (*vault.InitResult, error) {
2018-09-18 03:03:00 +00:00
ctx := namespace.ContextWithNamespace(context.Background(), namespace.RootNamespace)
var recoveryConfig *vault.SealConfig
barrierConfig := &vault.SealConfig{
SecretShares: 1,
SecretThreshold: 1,
}
if core.SealAccess().RecoveryKeySupported() {
recoveryConfig = &vault.SealConfig{
SecretShares: 1,
SecretThreshold: 1,
}
2015-03-31 23:44:47 +00:00
}
if core.SealAccess().StoredKeysSupported() != vaultseal.StoredKeysNotSupported {
barrierConfig.StoredShares = 1
}
2015-03-31 23:44:47 +00:00
// Initialize it with a basic single key
init, err := core.Initialize(ctx, &vault.InitParams{
BarrierConfig: barrierConfig,
RecoveryConfig: recoveryConfig,
})
2015-03-31 23:44:47 +00:00
if err != nil {
return nil, err
}
// Handle unseal with stored keys
if core.SealAccess().StoredKeysSupported() == vaultseal.StoredKeysSupportedGeneric {
err := core.UnsealWithStoredKeys(ctx)
if err != nil {
return nil, err
}
} else {
// Copy the key so that it can be zeroed
key := make([]byte, len(init.SecretShares[0]))
copy(key, init.SecretShares[0])
// Unseal the core
unsealed, err := core.Unseal(key)
if err != nil {
return nil, err
}
if !unsealed {
return nil, fmt.Errorf("failed to unseal Vault for dev mode")
}
2015-03-31 23:44:47 +00:00
}
isLeader, _, _, err := core.Leader()
if err != nil && err != vault.ErrHANotEnabled {
return nil, errwrap.Wrapf("failed to check active status: {{err}}", err)
}
if err == nil {
leaderCount := 5
for !isLeader {
if leaderCount == 0 {
buf := make([]byte, 1<<16)
runtime.Stack(buf, true)
return nil, fmt.Errorf("failed to get active status after five seconds; call stack is\n%s\n", buf)
}
time.Sleep(1 * time.Second)
isLeader, _, _, err = core.Leader()
if err != nil {
return nil, errwrap.Wrapf("failed to check active status: {{err}}", err)
}
leaderCount--
}
}
// Generate a dev root token if one is provided in the flag
if coreConfig.DevToken != "" {
req := &logical.Request{
2016-08-19 20:45:17 +00:00
ID: "dev-gen-root",
Operation: logical.UpdateOperation,
ClientToken: init.RootToken,
Path: "auth/token/create",
Data: map[string]interface{}{
"id": coreConfig.DevToken,
"policies": []string{"root"},
"no_parent": true,
"no_default_policy": true,
},
}
2018-09-18 03:03:00 +00:00
resp, err := core.HandleRequest(ctx, req)
if err != nil {
return nil, errwrap.Wrapf(fmt.Sprintf("failed to create root token with ID %q: {{err}}", coreConfig.DevToken), err)
}
if resp == nil {
return nil, fmt.Errorf("nil response when creating root token with ID %q", coreConfig.DevToken)
}
if resp.Auth == nil {
return nil, fmt.Errorf("nil auth when creating root token with ID %q", coreConfig.DevToken)
}
init.RootToken = resp.Auth.ClientToken
2016-08-19 20:45:17 +00:00
req.ID = "dev-revoke-init-root"
req.Path = "auth/token/revoke-self"
req.Data = nil
2018-09-18 03:03:00 +00:00
resp, err = core.HandleRequest(ctx, req)
if err != nil {
return nil, errwrap.Wrapf("failed to revoke initial root token: {{err}}", err)
}
}
2015-03-31 23:44:47 +00:00
// Set the token
if !c.flagDevNoStoreToken {
tokenHelper, err := c.TokenHelper()
if err != nil {
return nil, err
}
if err := tokenHelper.Store(init.RootToken); err != nil {
return nil, err
}
2015-03-31 23:44:47 +00:00
}
kvVer := "2"
if c.flagDevKVV1 || c.flagDevLeasedKV {
kvVer = "1"
}
req := &logical.Request{
Operation: logical.UpdateOperation,
ClientToken: init.RootToken,
Path: "sys/mounts/secret",
Data: map[string]interface{}{
"type": "kv",
"path": "secret/",
"description": "key/value secret storage",
"options": map[string]string{
"version": kvVer,
},
},
}
resp, err := core.HandleRequest(ctx, req)
if err != nil {
return nil, errwrap.Wrapf("error creating default K/V store: {{err}}", err)
}
if resp.IsError() {
return nil, errwrap.Wrapf("failed to create default K/V store: {{err}}", resp.Error())
}
2015-03-31 23:44:47 +00:00
return init, nil
}
func (c *ServerCommand) enableThreeNodeDevCluster(base *vault.CoreConfig, info map[string]string, infoKeys []string, devListenAddress, tempDir string) int {
testCluster := vault.NewTestCluster(&testing.RuntimeT{}, base, &vault.TestClusterOptions{
HandlerFunc: vaulthttp.Handler,
BaseListenAddress: c.flagDevListenAddr,
Logger: c.logger,
TempDir: tempDir,
})
defer c.cleanupGuard.Do(testCluster.Cleanup)
info["cluster parameters path"] = testCluster.TempDir
infoKeys = append(infoKeys, "cluster parameters path")
for i, core := range testCluster.Cores {
info[fmt.Sprintf("node %d api address", i)] = fmt.Sprintf("https://%s", core.Listeners[0].Address.String())
infoKeys = append(infoKeys, fmt.Sprintf("node %d api address", i))
}
infoKeys = append(infoKeys, "version")
verInfo := version.GetVersion()
info["version"] = verInfo.FullVersionNumber(false)
if verInfo.Revision != "" {
info["version sha"] = strings.Trim(verInfo.Revision, "'")
infoKeys = append(infoKeys, "version sha")
}
infoKeys = append(infoKeys, "cgo")
info["cgo"] = "disabled"
if version.CgoEnabled {
info["cgo"] = "enabled"
}
// Server configuration output
padding := 24
sort.Strings(infoKeys)
c.UI.Output("==> Vault server configuration:\n")
for _, k := range infoKeys {
c.UI.Output(fmt.Sprintf(
"%s%s: %s",
strings.Repeat(" ", padding-len(k)),
strings.Title(k),
info[k]))
}
c.UI.Output("")
for _, core := range testCluster.Cores {
core.Server.Handler = vaulthttp.Handler(&vault.HandlerProperties{
Core: core.Core,
})
core.SetClusterHandler(core.Server.Handler)
}
testCluster.Start()
2018-09-18 03:03:00 +00:00
ctx := namespace.ContextWithNamespace(context.Background(), namespace.RootNamespace)
if base.DevToken != "" {
req := &logical.Request{
ID: "dev-gen-root",
Operation: logical.UpdateOperation,
ClientToken: testCluster.RootToken,
Path: "auth/token/create",
Data: map[string]interface{}{
"id": base.DevToken,
"policies": []string{"root"},
"no_parent": true,
"no_default_policy": true,
},
}
2018-09-18 03:03:00 +00:00
resp, err := testCluster.Cores[0].HandleRequest(ctx, req)
if err != nil {
c.UI.Error(fmt.Sprintf("failed to create root token with ID %s: %s", base.DevToken, err))
return 1
}
if resp == nil {
c.UI.Error(fmt.Sprintf("nil response when creating root token with ID %s", base.DevToken))
return 1
}
if resp.Auth == nil {
c.UI.Error(fmt.Sprintf("nil auth when creating root token with ID %s", base.DevToken))
return 1
}
testCluster.RootToken = resp.Auth.ClientToken
req.ID = "dev-revoke-init-root"
req.Path = "auth/token/revoke-self"
req.Data = nil
2018-09-18 03:03:00 +00:00
resp, err = testCluster.Cores[0].HandleRequest(ctx, req)
if err != nil {
c.UI.Output(fmt.Sprintf("failed to revoke initial root token: %s", err))
return 1
}
}
// Set the token
tokenHelper, err := c.TokenHelper()
if err != nil {
c.UI.Error(fmt.Sprintf("Error getting token helper: %s", err))
return 1
}
if err := tokenHelper.Store(testCluster.RootToken); err != nil {
c.UI.Error(fmt.Sprintf("Error storing in token helper: %s", err))
return 1
}
if err := ioutil.WriteFile(filepath.Join(testCluster.TempDir, "root_token"), []byte(testCluster.RootToken), 0755); err != nil {
c.UI.Error(fmt.Sprintf("Error writing token to tempfile: %s", err))
return 1
}
c.UI.Output(fmt.Sprintf(
"==> Three node dev mode is enabled\n\n" +
"The unseal key and root token are reproduced below in case you\n" +
"want to seal/unseal the Vault or play with authentication.\n",
))
for i, key := range testCluster.BarrierKeys {
c.UI.Output(fmt.Sprintf(
"Unseal Key %d: %s",
i+1, base64.StdEncoding.EncodeToString(key),
))
}
c.UI.Output(fmt.Sprintf(
"\nRoot Token: %s\n", testCluster.RootToken,
))
c.UI.Output(fmt.Sprintf(
"\nUseful env vars:\n"+
"VAULT_TOKEN=%s\n"+
"VAULT_ADDR=%s\n"+
"VAULT_CACERT=%s/ca_cert.pem\n",
testCluster.RootToken,
testCluster.Cores[0].Client.Address(),
testCluster.TempDir,
))
// Output the header that the server has started
c.UI.Output("==> Vault server started! Log data will stream in below:\n")
// Inform any tests that the server is ready
select {
case c.startedCh <- struct{}{}:
default:
}
// Release the log gate.
c.logGate.Flush()
// Wait for shutdown
shutdownTriggered := false
for !shutdownTriggered {
select {
case <-c.ShutdownCh:
c.UI.Output("==> Vault shutdown triggered")
2018-03-20 18:54:10 +00:00
// Stop the listeners so that we don't process further client requests.
c.cleanupGuard.Do(testCluster.Cleanup)
// Shutdown will wait until after Vault is sealed, which means the
// request forwarding listeners will also be closed (and also
// waited for).
for _, core := range testCluster.Cores {
if err := core.Shutdown(); err != nil {
c.UI.Error(fmt.Sprintf("Error with core shutdown: %s", err))
}
}
shutdownTriggered = true
case <-c.SighupCh:
c.UI.Output("==> Vault reload triggered")
for _, core := range testCluster.Cores {
if err := c.Reload(core.ReloadFuncsLock, core.ReloadFuncs, nil); err != nil {
c.UI.Error(fmt.Sprintf("Error(s) were encountered during reload: %s", err))
}
}
}
}
return 0
}
// addPlugin adds any plugins to the catalog
func (c *ServerCommand) addPlugin(path, token string, core *vault.Core) error {
// Get the sha256 of the file at the given path.
pluginSum := func(p string) (string, error) {
hasher := sha256.New()
f, err := os.Open(p)
if err != nil {
return "", err
}
defer f.Close()
if _, err := io.Copy(hasher, f); err != nil {
return "", err
}
return hex.EncodeToString(hasher.Sum(nil)), nil
}
// Mount any test plugins. We do this explicitly before we inform tests of
// a completely booted server intentionally.
sha256sum, err := pluginSum(path)
if err != nil {
return err
}
// Default the name to the basename of the binary
name := filepath.Base(path)
// File a request against core to enable the plugin
req := &logical.Request{
Operation: logical.UpdateOperation,
ClientToken: token,
2018-11-07 01:21:24 +00:00
Path: fmt.Sprintf("sys/plugins/catalog/%s", name),
Data: map[string]interface{}{
"sha256": sha256sum,
"command": name,
},
}
2018-09-18 03:03:00 +00:00
ctx := namespace.ContextWithNamespace(context.Background(), namespace.RootNamespace)
if _, err := core.HandleRequest(ctx, req); err != nil {
return err
}
return nil
}
// detectRedirect is used to attempt redirect address detection
func (c *ServerCommand) detectRedirect(detect physical.RedirectDetect,
config *server.Config) (string, error) {
// Get the hostname
host, err := detect.DetectHostAddr()
if err != nil {
return "", err
}
2015-12-22 20:40:36 +00:00
// set [] for ipv6 addresses
if strings.Contains(host, ":") && !strings.Contains(host, "]") {
host = "[" + host + "]"
}
// Default the port and scheme
scheme := "https"
port := 8200
// Attempt to detect overrides
for _, list := range config.Listeners {
// Only attempt TCP
if list.Type != "tcp" {
continue
}
// Check if TLS is disabled
2015-05-15 20:41:30 +00:00
if val, ok := list.Config["tls_disable"]; ok {
disable, err := parseutil.ParseBool(val)
2015-05-15 20:41:30 +00:00
if err != nil {
return "", errwrap.Wrapf("tls_disable: {{err}}", err)
2015-05-15 20:41:30 +00:00
}
if disable {
scheme = "http"
}
}
// Check for address override
var addr string
addrRaw, ok := list.Config["address"]
if !ok {
addr = "127.0.0.1:8200"
} else {
addr = addrRaw.(string)
}
// Check for localhost
hostStr, portStr, err := net.SplitHostPort(addr)
if err != nil {
continue
}
if hostStr == "127.0.0.1" {
host = hostStr
}
// Check for custom port
listPort, err := strconv.Atoi(portStr)
if err != nil {
continue
}
port = listPort
}
// Build a URL
url := &url.URL{
Scheme: scheme,
Host: fmt.Sprintf("%s:%d", host, port),
}
// Return the URL string
return url.String(), nil
}
// setupTelemetry is used to setup the telemetry sub-systems and returns the in-memory sink to be used in http configuration
func (c *ServerCommand) setupTelemetry(config *server.Config) (*metricsutil.MetricsHelper, error) {
/* Setup telemetry
Aggregate on 10 second intervals for 1 minute. Expose the
metrics over stderr when there is a SIGUSR1 received.
*/
inm := metrics.NewInmemSink(10*time.Second, time.Minute)
metrics.DefaultInmemSignal(inm)
2015-07-14 22:27:18 +00:00
var telConfig *server.Telemetry
if config.Telemetry != nil {
2015-07-14 22:27:18 +00:00
telConfig = config.Telemetry
} else {
telConfig = &server.Telemetry{}
2015-07-14 22:27:18 +00:00
}
metricsConf := metrics.DefaultConfig("vault")
2015-07-14 22:27:18 +00:00
metricsConf.EnableHostname = !telConfig.DisableHostname
metricsConf.EnableHostnameLabel = telConfig.EnableHostnameLabel
// Configure the statsite sink
var fanout metrics.FanoutSink
var prometheusEnabled bool
// Configure the Prometheus sink
if telConfig.PrometheusRetentionTime != 0 {
prometheusEnabled = true
prometheusOpts := prometheus.PrometheusOpts{
Expiration: telConfig.PrometheusRetentionTime,
}
sink, err := prometheus.NewPrometheusSinkFrom(prometheusOpts)
if err != nil {
return nil, err
}
fanout = append(fanout, sink)
}
metricHelper := metricsutil.NewMetricsHelper(inm, prometheusEnabled)
2015-07-14 22:27:18 +00:00
if telConfig.StatsiteAddr != "" {
sink, err := metrics.NewStatsiteSink(telConfig.StatsiteAddr)
if err != nil {
return nil, err
}
fanout = append(fanout, sink)
}
// Configure the statsd sink
2015-07-14 22:27:18 +00:00
if telConfig.StatsdAddr != "" {
sink, err := metrics.NewStatsdSink(telConfig.StatsdAddr)
if err != nil {
return nil, err
}
fanout = append(fanout, sink)
}
// Configure the Circonus sink
if telConfig.CirconusAPIToken != "" || telConfig.CirconusCheckSubmissionURL != "" {
cfg := &circonus.Config{}
cfg.Interval = telConfig.CirconusSubmissionInterval
cfg.CheckManager.API.TokenKey = telConfig.CirconusAPIToken
cfg.CheckManager.API.TokenApp = telConfig.CirconusAPIApp
cfg.CheckManager.API.URL = telConfig.CirconusAPIURL
cfg.CheckManager.Check.SubmissionURL = telConfig.CirconusCheckSubmissionURL
cfg.CheckManager.Check.ID = telConfig.CirconusCheckID
cfg.CheckManager.Check.ForceMetricActivation = telConfig.CirconusCheckForceMetricActivation
cfg.CheckManager.Check.InstanceID = telConfig.CirconusCheckInstanceID
cfg.CheckManager.Check.SearchTag = telConfig.CirconusCheckSearchTag
2016-11-10 21:17:55 +00:00
cfg.CheckManager.Check.DisplayName = telConfig.CirconusCheckDisplayName
cfg.CheckManager.Check.Tags = telConfig.CirconusCheckTags
cfg.CheckManager.Broker.ID = telConfig.CirconusBrokerID
cfg.CheckManager.Broker.SelectTag = telConfig.CirconusBrokerSelectTag
if cfg.CheckManager.API.TokenApp == "" {
cfg.CheckManager.API.TokenApp = "vault"
}
if cfg.CheckManager.Check.DisplayName == "" {
cfg.CheckManager.Check.DisplayName = "Vault"
}
if cfg.CheckManager.Check.SearchTag == "" {
cfg.CheckManager.Check.SearchTag = "service:vault"
}
sink, err := circonus.NewCirconusSink(cfg)
if err != nil {
return nil, err
}
sink.Start()
fanout = append(fanout, sink)
}
if telConfig.DogStatsDAddr != "" {
var tags []string
if telConfig.DogStatsDTags != nil {
tags = telConfig.DogStatsDTags
}
sink, err := datadog.NewDogStatsdSink(telConfig.DogStatsDAddr, metricsConf.HostName)
if err != nil {
return nil, errwrap.Wrapf("failed to start DogStatsD sink: {{err}}", err)
}
sink.SetTags(tags)
fanout = append(fanout, sink)
}
// Configure the stackdriver sink
if telConfig.StackdriverProjectID != "" {
client, err := monitoring.NewMetricClient(context.Background(), option.WithUserAgent(useragent.String()))
if err != nil {
return nil, fmt.Errorf("Failed to create stackdriver client: %v", err)
}
sink := stackdriver.NewSink(client, &stackdriver.Config{
ProjectID: telConfig.StackdriverProjectID,
Location: telConfig.StackdriverLocation,
Namespace: telConfig.StackdriverNamespace,
})
fanout = append(fanout, sink)
}
// Initialize the global sink
if len(fanout) > 1 {
// Hostname enabled will create poor quality metrics name for prometheus
if !telConfig.DisableHostname {
c.UI.Warn("telemetry.disable_hostname has been set to false. Recommended setting is true for Prometheus to avoid poorly named metrics.")
}
} else {
metricsConf.EnableHostname = false
}
fanout = append(fanout, inm)
_, err := metrics.NewGlobal(metricsConf, fanout)
if err != nil {
return nil, err
}
return metricHelper, nil
}
func (c *ServerCommand) Reload(lock *sync.RWMutex, reloadFuncs *map[string][]reload.ReloadFunc, configPath []string) error {
lock.RLock()
defer lock.RUnlock()
var reloadErrors *multierror.Error
for k, relFuncs := range *reloadFuncs {
switch {
case strings.HasPrefix(k, "listener|"):
for _, relFunc := range relFuncs {
if relFunc != nil {
if err := relFunc(nil); err != nil {
reloadErrors = multierror.Append(reloadErrors, errwrap.Wrapf("error encountered reloading listener: {{err}}", err))
}
}
}
case strings.HasPrefix(k, "audit_file|"):
for _, relFunc := range relFuncs {
if relFunc != nil {
if err := relFunc(nil); err != nil {
reloadErrors = multierror.Append(reloadErrors, errwrap.Wrapf(fmt.Sprintf("error encountered reloading file audit device at path %q: {{err}}", strings.TrimPrefix(k, "audit_file|")), err))
}
}
}
}
}
// Send a message that we reloaded. This prevents "guessing" sleep times
// in tests.
select {
case c.reloadedCh <- struct{}{}:
default:
}
return reloadErrors.ErrorOrNil()
}
// storePidFile is used to write out our PID to a file if necessary
func (c *ServerCommand) storePidFile(pidPath string) error {
// Quit fast if no pidfile
if pidPath == "" {
return nil
}
// Open the PID file
pidFile, err := os.OpenFile(pidPath, os.O_CREATE|os.O_WRONLY|os.O_TRUNC, 0644)
if err != nil {
return errwrap.Wrapf("could not open pid file: {{err}}", err)
}
defer pidFile.Close()
// Write out the PID
pid := os.Getpid()
_, err = pidFile.WriteString(fmt.Sprintf("%d", pid))
if err != nil {
return errwrap.Wrapf("could not write to pid file: {{err}}", err)
}
return nil
}
// removePidFile is used to cleanup the PID file if necessary
func (c *ServerCommand) removePidFile(pidPath string) error {
if pidPath == "" {
return nil
}
return os.Remove(pidPath)
}
2018-10-23 06:34:02 +00:00
// storageMigrationActive checks and warns against in-progress storage migrations.
// This function will block until storage is available.
2018-10-23 06:34:02 +00:00
func (c *ServerCommand) storageMigrationActive(backend physical.Backend) bool {
first := true
for {
2018-10-23 06:34:02 +00:00
migrationStatus, err := CheckStorageMigration(backend)
if err == nil {
if migrationStatus != nil {
startTime := migrationStatus.Start.Format(time.RFC3339)
c.UI.Error(wrapAtLength(fmt.Sprintf("ERROR! Storage migration in progress (started: %s). "+
"Server startup is prevented until the migration completes. Use 'vault operator migrate -reset' "+
"to force clear the migration lock.", startTime)))
return true
}
return false
}
if first {
first = false
2018-10-23 06:34:02 +00:00
c.UI.Warn("\nWARNING! Unable to read storage migration status.")
// unexpected state, so stop buffering log messages
c.logGate.Flush()
}
2018-10-23 06:34:02 +00:00
c.logger.Warn("storage migration check error", "error", err.Error())
select {
case <-time.After(2 * time.Second):
case <-c.ShutdownCh:
return true
}
}
}
2018-10-23 06:34:02 +00:00
type StorageMigrationStatus struct {
Start time.Time `json:"start"`
}
2018-10-23 06:34:02 +00:00
func CheckStorageMigration(b physical.Backend) (*StorageMigrationStatus, error) {
entry, err := b.Get(context.Background(), storageMigrationLock)
if err != nil {
return nil, err
}
if entry == nil {
return nil, nil
}
2018-10-23 06:34:02 +00:00
var status StorageMigrationStatus
if err := jsonutil.DecodeJSON(entry.Value, &status); err != nil {
return nil, err
}
return &status, nil
}
2018-10-23 06:34:02 +00:00
func SetStorageMigration(b physical.Backend, active bool) error {
if !active {
2018-10-23 06:34:02 +00:00
return b.Delete(context.Background(), storageMigrationLock)
}
2018-10-23 06:34:02 +00:00
status := StorageMigrationStatus{
Start: time.Now(),
}
enc, err := jsonutil.EncodeJSON(status)
if err != nil {
return err
}
entry := &physical.Entry{
2018-10-23 06:34:02 +00:00
Key: storageMigrationLock,
Value: enc,
}
return b.Put(context.Background(), entry)
}
2016-08-19 20:45:17 +00:00
type grpclogFaker struct {
logger log.Logger
2017-10-10 16:27:51 +00:00
log bool
2016-08-19 20:45:17 +00:00
}
func (g *grpclogFaker) Fatal(args ...interface{}) {
g.logger.Error(fmt.Sprint(args...))
os.Exit(1)
}
func (g *grpclogFaker) Fatalf(format string, args ...interface{}) {
g.logger.Error(fmt.Sprintf(format, args...))
os.Exit(1)
}
func (g *grpclogFaker) Fatalln(args ...interface{}) {
g.logger.Error(fmt.Sprintln(args...))
os.Exit(1)
}
func (g *grpclogFaker) Print(args ...interface{}) {
if g.log && g.logger.IsDebug() {
g.logger.Debug(fmt.Sprint(args...))
2017-10-10 16:27:51 +00:00
}
2016-08-19 20:45:17 +00:00
}
func (g *grpclogFaker) Printf(format string, args ...interface{}) {
if g.log && g.logger.IsDebug() {
g.logger.Debug(fmt.Sprintf(format, args...))
2017-10-10 16:27:51 +00:00
}
2016-08-19 20:45:17 +00:00
}
func (g *grpclogFaker) Println(args ...interface{}) {
if g.log && g.logger.IsDebug() {
g.logger.Debug(fmt.Sprintln(args...))
2017-10-10 16:27:51 +00:00
}
2016-08-19 20:45:17 +00:00
}