feat(kms): AppRole login with background token renewal and fail-closed expiry (#5487)

* feat(kms): add AppRole configuration surface for Vault auth

Extend VaultAuthMethod::AppRole with secret_id_file (re-read on every
login so external rotation is picked up), a configurable auth mount
(default "approle"), and an optional fail-closed safety window. All new
fields are serde(default) so previously persisted configurations keep
deserializing, and the strict admin-configure deserializer accepts them
as optional.

Environment selection: setting RUSTFS_KMS_VAULT_APPROLE_ROLE_ID switches
both Vault backends to AppRole; the secret_id comes from
RUSTFS_KMS_VAULT_APPROLE_SECRET_ID_FILE (path stored, file wins) or
RUSTFS_KMS_VAULT_APPROLE_SECRET_ID, following the static secret-key file
precedent. validate() rejects AppRole configs without a role_id, without
any secret_id source, or with an empty mount.

Also append the CredentialsUnavailable error variant used by the
fail-closed credential gate.

* feat(kms): implement AppRole login with background renewal and fail-closed expiry

Implement the AppRoleLogin token source (vaultrs approle login +
renew-self) and wire lease-bound credentials through the provider:

- Each successful login/renewal installs a new client generation in the
  ArcSwap; in-flight requests finish on the generation they captured.
- A background renewal task refreshes at half the lease TTL: renewable
  tokens are renewed in place, everything else (or a failed renewal)
  falls back to a fresh login. Auth exchanges run under the typed retry
  policy (OpClass::Auth) and failed cycles retry on a fixed cadence, so
  the provider recovers once Vault does.
- Fail-closed: current() refuses to hand out a token inside the
  configured safety window of its expiry (default: one attempt timeout),
  returning CredentialsUnavailable instead of sending a request whose
  token may lapse mid-flight.
- Refreshes are single-flight: concurrent triggers for the same
  generation coalesce into one login.
- The renewal task's owner handle lives on the KMS service version:
  stop() shuts it down explicitly and reconfigure recycles it via
  cancel-on-drop when the old version is discarded.
- The secret_id file is re-read on every login attempt; missing or empty
  files fail the attempt without contacting Vault. Crate-owned copies of
  tokens and secret_ids are zeroized on drop, and Debug output of every
  credential-carrying type stays redacted (leak regression tests).

The renewal machinery is covered by paused-clock tests driving a
scripted token source: renew-at-half-TTL timing, login fallback,
fail-closed window entry and recovery, prompt task recycling, and
coalesced concurrent refreshes.

* feat(kms): add Vault Agent token file authentication

Add the TokenFile source: the token is read from an agent-managed sink
file (RUSTFS_KMS_VAULT_TOKEN_FILE or the TokenFile auth config) and
re-read once per poll interval (default 30s) through the existing
renewal loop, so a token rotated by the agent installs a new client
generation within one poll of the atomic replace. Each successful read
extends the token's observed validity to twice the poll interval; a
file that disappears or turns empty keeps failing the refresh until the
fail-closed window trips, and heals the provider as soon as it is
restored.

Reads are strict and never contact Vault on failure: the file must be
non-empty after trimming, and on Unix group/other permission bits are a
hard error (mirroring the SFTP host-key rule). Rotation detection uses
a content digest; the token itself is never stored on the source and
the crate-owned copy is zeroized.

Configuring the token file together with AppRole or an explicit static
token is rejected as a configuration error. All new config fields are
serde(default) and the strict admin-configure deserializer accepts the
new variant.

Covered by paused-clock tests (atomic replacement installs a new
generation next cycle, deletion fails closed and recovers, prompt task
recycling) plus negatives for missing/empty/over-permissive files and a
Debug leak regression.

* docs(kms): add Vault authentication and credential lifecycle runbook

Cover choosing between static token, AppRole, and Vault Agent token
file auth; AppRole role setup with SecretID delivery and rotation;
Agent sink deployment with the permission requirements; and the
fail-closed window semantics with a troubleshooting table keyed on the
renewal task's log lines.
This commit is contained in:
Zhengchao An
2026-07-31 06:29:49 +08:00
committed by GitHub
parent 342ee1df78
commit 3921336b23
9 changed files with 1813 additions and 147 deletions
+401 -8
View File
@@ -29,8 +29,14 @@ pub const ENV_KMS_VAULT_TRANSIT_METADATA_KV_MOUNT: &str = "RUSTFS_KMS_VAULT_TRAN
pub const ENV_KMS_VAULT_TRANSIT_METADATA_PREFIX: &str = "RUSTFS_KMS_VAULT_TRANSIT_METADATA_PREFIX";
pub const ENV_KMS_STATIC_SECRET_KEY: &str = "RUSTFS_KMS_STATIC_SECRET_KEY";
pub const ENV_KMS_STATIC_SECRET_KEY_FILE: &str = "RUSTFS_KMS_STATIC_SECRET_KEY_FILE";
pub const ENV_KMS_VAULT_APPROLE_ROLE_ID: &str = "RUSTFS_KMS_VAULT_APPROLE_ROLE_ID";
pub const ENV_KMS_VAULT_APPROLE_SECRET_ID: &str = "RUSTFS_KMS_VAULT_APPROLE_SECRET_ID";
pub const ENV_KMS_VAULT_APPROLE_SECRET_ID_FILE: &str = "RUSTFS_KMS_VAULT_APPROLE_SECRET_ID_FILE";
pub const ENV_KMS_VAULT_APPROLE_MOUNT: &str = "RUSTFS_KMS_VAULT_APPROLE_MOUNT";
pub const ENV_KMS_VAULT_TOKEN_FILE: &str = "RUSTFS_KMS_VAULT_TOKEN_FILE";
pub const DEFAULT_VAULT_TRANSIT_METADATA_KV_MOUNT: &str = "secret";
pub const DEFAULT_VAULT_TRANSIT_METADATA_KEY_PREFIX: &str = "rustfs/kms/transit-metadata";
pub const DEFAULT_VAULT_APPROLE_MOUNT: &str = "approle";
/// Upper bound applied to `KmsConfig::timeout` when deriving backend behavior.
///
@@ -53,6 +59,10 @@ fn default_vault_kv2_mount_path() -> String {
"transit".to_string()
}
fn default_vault_approle_mount() -> String {
DEFAULT_VAULT_APPROLE_MOUNT.to_string()
}
pub const KMS_CONFIG_REDACTION_RULES: &[RedactionRule] = &[
RedactionRule::new("kms.local.master_key", RedactionLevel::Secret, "local backend key encryption material"),
RedactionRule::new("kms.vault.token", RedactionLevel::Secret, "vault authentication token"),
@@ -388,18 +398,94 @@ impl Default for VaultTransitConfig {
pub enum VaultAuthMethod {
/// Token authentication
Token { token: String },
/// AppRole authentication
AppRole { role_id: String, secret_id: String },
/// AppRole authentication: login with `role_id` + `secret_id` for a
/// lease-bound token that is renewed in the background.
AppRole {
role_id: String,
/// Inline secret_id; used only when `secret_id_file` is unset.
secret_id: String,
/// Path to a file holding the secret_id. Re-read on every login so an
/// externally rotated secret_id is picked up; takes precedence over the
/// inline value.
#[serde(default)]
secret_id_file: Option<PathBuf>,
/// AppRole auth engine mount path.
#[serde(default = "default_vault_approle_mount")]
mount: String,
/// Fail-closed margin in seconds: once the current token is within this
/// window of expiry without a successful refresh, requests are refused
/// instead of sent with a token that may lapse mid-flight. Defaults to
/// the per-attempt timeout.
#[serde(default)]
refresh_safety_window_secs: Option<u64>,
},
/// Agent-managed token file (for example a Vault Agent auto-auth sink):
/// the token is read from `path` and re-read periodically so a token
/// rotated by the agent is picked up without a restart.
TokenFile {
path: PathBuf,
/// Seconds between token file re-reads. Each successful read also
/// extends the token's observed validity to twice this value, so a
/// file that stops being readable eventually trips the fail-closed
/// window. Defaults to 30 seconds.
#[serde(default)]
poll_interval_secs: Option<u64>,
/// Fail-closed margin in seconds, as on `AppRole`. Defaults to the
/// per-attempt timeout.
#[serde(default)]
refresh_safety_window_secs: Option<u64>,
},
}
impl VaultAuthMethod {
/// AppRole authentication with the default mount and no secret-id file.
pub fn approle(role_id: String, secret_id: String) -> Self {
Self::AppRole {
role_id,
secret_id,
secret_id_file: None,
mount: default_vault_approle_mount(),
refresh_safety_window_secs: None,
}
}
/// Agent-managed token file with the default poll interval.
pub fn token_file(path: PathBuf) -> Self {
Self::TokenFile {
path,
poll_interval_secs: None,
refresh_safety_window_secs: None,
}
}
}
impl fmt::Debug for VaultAuthMethod {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match self {
Self::Token { token } => f.debug_struct("Token").field("token", &redacted_secret(token)).finish(),
Self::AppRole { role_id, secret_id } => f
Self::AppRole {
role_id,
secret_id,
secret_id_file,
mount,
refresh_safety_window_secs,
} => f
.debug_struct("AppRole")
.field("role_id", role_id)
.field("secret_id", &redacted_secret(secret_id))
.field("secret_id_file", secret_id_file)
.field("mount", mount)
.field("refresh_safety_window_secs", refresh_safety_window_secs)
.finish(),
Self::TokenFile {
path,
poll_interval_secs,
refresh_safety_window_secs,
} => f
.debug_struct("TokenFile")
.field("path", path)
.field("poll_interval_secs", poll_interval_secs)
.field("refresh_safety_window_secs", refresh_safety_window_secs)
.finish(),
}
}
@@ -479,7 +565,7 @@ impl KmsConfig {
backend: KmsBackend::VaultKv2,
backend_config: BackendConfig::VaultKv2(Box::new(VaultConfig {
address: address.to_string(),
auth_method: VaultAuthMethod::AppRole { role_id, secret_id },
auth_method: VaultAuthMethod::approle(role_id, secret_id),
..Default::default()
})),
..Default::default()
@@ -637,6 +723,8 @@ impl KmsConfig {
return Err(KmsError::configuration_error("Vault KV2 address must use http or https scheme"));
}
validate_vault_auth_method("Vault KV2", &config.auth_method)?;
if !self.allow_insecure_dev_defaults {
validate_vault_development_defaults("Vault KV2", &config.address, &config.auth_method, config.tls.as_ref())?;
}
@@ -660,6 +748,8 @@ impl KmsConfig {
return Err(KmsError::configuration_error("Vault Transit address must use http or https scheme"));
}
validate_vault_auth_method("Vault Transit", &config.auth_method)?;
if !self.allow_insecure_dev_defaults {
validate_vault_development_defaults(
"Vault Transit",
@@ -764,7 +854,7 @@ impl KmsConfig {
}
KmsBackend::VaultKv2 => {
let address = get_env_str("RUSTFS_KMS_VAULT_ADDRESS", "http://localhost:8200");
let token = get_env_str("RUSTFS_KMS_VAULT_TOKEN", "dev-token");
let auth_method = vault_auth_method_from_env()?;
let skip_tls_verify = get_env_bool(ENV_KMS_VAULT_SKIP_TLS_VERIFY, false);
let mount_path = match get_env_opt_str("RUSTFS_KMS_VAULT_MOUNT_PATH") {
@@ -779,7 +869,7 @@ impl KmsConfig {
config.backend_config = BackendConfig::VaultKv2(Box::new(VaultConfig {
address,
auth_method: VaultAuthMethod::Token { token },
auth_method,
namespace: get_env_opt_str("RUSTFS_KMS_VAULT_NAMESPACE"),
mount_path,
kv_mount: get_env_str("RUSTFS_KMS_VAULT_KV_MOUNT", "secret"),
@@ -789,12 +879,12 @@ impl KmsConfig {
}
KmsBackend::VaultTransit => {
let address = get_env_str("RUSTFS_KMS_VAULT_ADDRESS", "http://localhost:8200");
let token = get_env_str("RUSTFS_KMS_VAULT_TOKEN", "dev-token");
let auth_method = vault_auth_method_from_env()?;
let skip_tls_verify = get_env_bool(ENV_KMS_VAULT_SKIP_TLS_VERIFY, false);
config.backend_config = BackendConfig::VaultTransit(Box::new(VaultTransitConfig {
address,
auth_method: VaultAuthMethod::Token { token },
auth_method,
namespace: get_env_opt_str("RUSTFS_KMS_VAULT_NAMESPACE"),
mount_path: get_env_str("RUSTFS_KMS_VAULT_MOUNT_PATH", "transit"),
metadata_kv_mount: get_env_str(
@@ -873,6 +963,95 @@ fn is_under_temp_dir(path: &Path) -> bool {
path.starts_with(std::env::temp_dir())
}
/// Resolve the Vault auth method from environment variables.
///
/// Setting `RUSTFS_KMS_VAULT_APPROLE_ROLE_ID` selects AppRole authentication;
/// the secret_id then comes from `RUSTFS_KMS_VAULT_APPROLE_SECRET_ID_FILE`
/// (re-read on every login, mirroring the `RUSTFS_KMS_STATIC_SECRET_KEY_FILE`
/// precedent) or inline from `RUSTFS_KMS_VAULT_APPROLE_SECRET_ID`, with the
/// file taking precedence. Without a role id the legacy token flow applies.
fn vault_auth_method_from_env() -> Result<VaultAuthMethod> {
if let Some(token_file) = get_env_opt_str(ENV_KMS_VAULT_TOKEN_FILE) {
// A token file names one authoritative credential source; combining it
// with another one would leave the effective identity ambiguous, so
// that is a configuration error rather than a precedence rule.
if get_env_opt_str(ENV_KMS_VAULT_APPROLE_ROLE_ID).is_some() {
return Err(KmsError::configuration_error(format!(
"{ENV_KMS_VAULT_TOKEN_FILE} cannot be combined with {ENV_KMS_VAULT_APPROLE_ROLE_ID}; configure exactly one Vault auth method"
)));
}
if get_env_opt_str("RUSTFS_KMS_VAULT_TOKEN").is_some() {
return Err(KmsError::configuration_error(format!(
"{ENV_KMS_VAULT_TOKEN_FILE} cannot be combined with RUSTFS_KMS_VAULT_TOKEN; configure exactly one Vault auth method"
)));
}
return Ok(VaultAuthMethod::token_file(PathBuf::from(token_file)));
}
let Some(role_id) = get_env_opt_str(ENV_KMS_VAULT_APPROLE_ROLE_ID) else {
return Ok(VaultAuthMethod::Token {
token: get_env_str("RUSTFS_KMS_VAULT_TOKEN", "dev-token"),
});
};
let secret_id_file = get_env_opt_str(ENV_KMS_VAULT_APPROLE_SECRET_ID_FILE).map(PathBuf::from);
let secret_id = get_env_opt_str(ENV_KMS_VAULT_APPROLE_SECRET_ID).unwrap_or_default();
if secret_id.is_empty() && secret_id_file.is_none() {
return Err(KmsError::configuration_error(format!(
"Vault AppRole requires {ENV_KMS_VAULT_APPROLE_SECRET_ID} or {ENV_KMS_VAULT_APPROLE_SECRET_ID_FILE} to be set"
)));
}
Ok(VaultAuthMethod::AppRole {
role_id,
secret_id,
secret_id_file,
mount: get_env_str(ENV_KMS_VAULT_APPROLE_MOUNT, DEFAULT_VAULT_APPROLE_MOUNT),
refresh_safety_window_secs: None,
})
}
fn validate_vault_auth_method(backend_name: &str, auth_method: &VaultAuthMethod) -> Result<()> {
match auth_method {
VaultAuthMethod::Token { .. } => Ok(()),
VaultAuthMethod::AppRole {
role_id,
secret_id,
secret_id_file,
mount,
..
} => {
if role_id.is_empty() {
return Err(KmsError::configuration_error(format!("{backend_name} AppRole role_id cannot be empty")));
}
if secret_id.is_empty() && secret_id_file.is_none() {
return Err(KmsError::configuration_error(format!(
"{backend_name} AppRole requires a secret_id or a secret_id_file"
)));
}
if mount.is_empty() {
return Err(KmsError::configuration_error(format!("{backend_name} AppRole mount cannot be empty")));
}
Ok(())
}
VaultAuthMethod::TokenFile {
path,
poll_interval_secs,
..
} => {
if path.as_os_str().is_empty() {
return Err(KmsError::configuration_error(format!("{backend_name} token file path cannot be empty")));
}
if poll_interval_secs == &Some(0) {
return Err(KmsError::configuration_error(format!(
"{backend_name} token file poll interval must be greater than 0"
)));
}
Ok(())
}
}
}
fn validate_vault_development_defaults(
backend_name: &str,
address: &str,
@@ -1297,6 +1476,220 @@ mod tests {
);
}
#[test]
fn test_from_env_selects_approle_when_role_id_is_set() {
with_vars(
vec![
("RUSTFS_KMS_BACKEND", Some("vault")),
("RUSTFS_KMS_VAULT_ADDRESS", Some("https://vault.example.com")),
(ENV_KMS_VAULT_APPROLE_ROLE_ID, Some("env-role-id")),
(ENV_KMS_VAULT_APPROLE_SECRET_ID, Some("env-approle-secret-id")),
(ENV_KMS_VAULT_APPROLE_MOUNT, Some("approle-alt")),
// A stale token env var must not override the AppRole selection.
("RUSTFS_KMS_VAULT_TOKEN", Some("vault-token")),
],
|| {
let config = KmsConfig::from_env().expect("kms config should load from env");
let vault = config.vault_config().expect("vault backend config");
let VaultAuthMethod::AppRole {
role_id,
secret_id,
secret_id_file,
mount,
refresh_safety_window_secs,
} = &vault.auth_method
else {
panic!("role id in the environment must select AppRole auth, got {:?}", vault.auth_method);
};
assert_eq!(role_id, "env-role-id");
assert_eq!(secret_id, "env-approle-secret-id");
assert_eq!(secret_id_file, &None);
assert_eq!(mount, "approle-alt");
assert_eq!(refresh_safety_window_secs, &None);
},
);
}
#[test]
fn test_from_env_approle_secret_id_file_is_stored_as_path() {
with_vars(
vec![
("RUSTFS_KMS_BACKEND", Some("vault-transit")),
("RUSTFS_KMS_VAULT_ADDRESS", Some("https://vault.example.com")),
(ENV_KMS_VAULT_APPROLE_ROLE_ID, Some("env-role-id")),
(ENV_KMS_VAULT_APPROLE_SECRET_ID_FILE, Some("/etc/rustfs/approle-secret-id")),
],
|| {
let config = KmsConfig::from_env().expect("kms config should load from env");
let vault = config.vault_transit_config().expect("vault transit backend config");
let VaultAuthMethod::AppRole {
secret_id,
secret_id_file,
mount,
..
} = &vault.auth_method
else {
panic!("role id in the environment must select AppRole auth");
};
// The path is stored, not read: the secret_id file is re-read on
// every login so external rotation is picked up.
assert_eq!(secret_id_file.as_deref(), Some(std::path::Path::new("/etc/rustfs/approle-secret-id")));
assert!(secret_id.is_empty());
assert_eq!(mount, DEFAULT_VAULT_APPROLE_MOUNT);
},
);
}
#[test]
fn test_from_env_approle_requires_secret_id_or_file() {
with_vars(
vec![
("RUSTFS_KMS_BACKEND", Some("vault")),
(ENV_KMS_VAULT_APPROLE_ROLE_ID, Some("env-role-id")),
(ENV_KMS_VAULT_APPROLE_SECRET_ID, None::<&str>),
(ENV_KMS_VAULT_APPROLE_SECRET_ID_FILE, None::<&str>),
],
|| {
let error = KmsConfig::from_env().expect_err("approle without a secret_id source must be rejected");
assert!(error.to_string().contains(ENV_KMS_VAULT_APPROLE_SECRET_ID));
},
);
}
#[test]
fn test_from_env_selects_token_file() {
with_vars(
vec![
("RUSTFS_KMS_BACKEND", Some("vault")),
("RUSTFS_KMS_VAULT_ADDRESS", Some("https://vault.example.com")),
(ENV_KMS_VAULT_TOKEN_FILE, Some("/run/vault-agent/token")),
],
|| {
let config = KmsConfig::from_env().expect("kms config should load from env");
let vault = config.vault_config().expect("vault backend config");
let VaultAuthMethod::TokenFile {
path,
poll_interval_secs,
refresh_safety_window_secs,
} = &vault.auth_method
else {
panic!("token file in the environment must select TokenFile auth, got {:?}", vault.auth_method);
};
assert_eq!(path, std::path::Path::new("/run/vault-agent/token"));
assert_eq!(poll_interval_secs, &None);
assert_eq!(refresh_safety_window_secs, &None);
},
);
}
#[test]
fn test_from_env_token_file_is_mutually_exclusive_with_other_auth() {
with_vars(
vec![
("RUSTFS_KMS_BACKEND", Some("vault")),
(ENV_KMS_VAULT_TOKEN_FILE, Some("/run/vault-agent/token")),
(ENV_KMS_VAULT_APPROLE_ROLE_ID, Some("env-role-id")),
],
|| {
let error = KmsConfig::from_env().expect_err("token file combined with approle must be rejected");
assert!(error.to_string().contains(ENV_KMS_VAULT_TOKEN_FILE));
assert!(error.to_string().contains(ENV_KMS_VAULT_APPROLE_ROLE_ID));
},
);
with_vars(
vec![
("RUSTFS_KMS_BACKEND", Some("vault-transit")),
(ENV_KMS_VAULT_TOKEN_FILE, Some("/run/vault-agent/token")),
("RUSTFS_KMS_VAULT_TOKEN", Some("vault-token")),
],
|| {
let error = KmsConfig::from_env().expect_err("token file combined with a static token must be rejected");
assert!(error.to_string().contains(ENV_KMS_VAULT_TOKEN_FILE));
assert!(error.to_string().contains("RUSTFS_KMS_VAULT_TOKEN"));
},
);
}
#[test]
fn test_validate_rejects_bad_token_file_settings() {
let vault_config = |auth_method: VaultAuthMethod| KmsConfig {
backend: KmsBackend::VaultKv2,
backend_config: BackendConfig::VaultKv2(Box::new(VaultConfig {
address: "https://vault.example.com:8200".to_string(),
auth_method,
..Default::default()
})),
..Default::default()
};
let error = vault_config(VaultAuthMethod::token_file(PathBuf::new()))
.validate()
.expect_err("empty token file path must be rejected");
assert!(error.to_string().contains("path"));
let error = vault_config(VaultAuthMethod::TokenFile {
path: PathBuf::from("/run/vault-agent/token"),
poll_interval_secs: Some(0),
refresh_safety_window_secs: None,
})
.validate()
.expect_err("zero poll interval must be rejected");
assert!(error.to_string().contains("poll interval"));
vault_config(VaultAuthMethod::token_file(PathBuf::from("/run/vault-agent/token")))
.validate()
.expect("well-formed token file auth must validate");
}
#[test]
fn test_approle_config_deserializes_legacy_shape_with_defaults() {
// Persisted configurations from before the AppRole implementation only
// carry role_id and secret_id; the new fields must fill with defaults.
let legacy = serde_json::json!({
"AppRole": {
"role_id": "legacy-role",
"secret_id": "legacy-secret-id",
}
});
let auth: VaultAuthMethod = serde_json::from_value(legacy).expect("legacy AppRole config must keep deserializing");
let VaultAuthMethod::AppRole {
role_id,
secret_id_file,
mount,
refresh_safety_window_secs,
..
} = auth
else {
panic!("expected AppRole");
};
assert_eq!(role_id, "legacy-role");
assert_eq!(secret_id_file, None);
assert_eq!(mount, DEFAULT_VAULT_APPROLE_MOUNT);
assert_eq!(refresh_safety_window_secs, None);
}
#[test]
fn test_validate_rejects_incomplete_approle() {
let mut config = KmsConfig::vault_approle(
Url::parse("https://vault.example.com:8200").expect("vault URL"),
String::new(),
"secret-id".to_string(),
);
let error = config.validate().expect_err("empty role_id must be rejected");
assert!(error.to_string().contains("role_id"));
config = KmsConfig::vault_approle(
Url::parse("https://vault.example.com:8200").expect("vault URL"),
"role-id".to_string(),
String::new(),
);
let error = config
.validate()
.expect_err("approle without secret_id or secret_id_file must be rejected");
assert!(error.to_string().contains("secret_id"));
}
#[test]
fn test_from_env_requires_vault_development_opt_in() {
with_vars(