gsadmin/rustfs
1
0
Fork 0
mirror of https://github.com/rustfs/rustfs.git synced 2026-09-04 03:05:39 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(lifecycle): prevent eager date-expiry deletion on config update (#2708)

Browse Source
This commit is contained in:
houseme
2026-04-28 18:26:14 +08:00
committed by GitHub
parent e0b8c4fd42
commit 2953558f41
28 changed files with 3069 additions and 593 deletions
Download Patch File Download Diff File Expand all files Collapse all files
crates/ecstore/src/disk/disk_store.rs
+69 -5
View File
@@ -46,9 +46,7 @@ const DISK_HEALTH_FAULTY: u32 = 1;
pub const ENV_RUSTFS_DRIVE_ACTIVE_MONITORING: &str = "RUSTFS_DRIVE_ACTIVE_MONITORING";
pub const DEFAULT_RUSTFS_DRIVE_ACTIVE_MONITORING: bool = true;
pub const CHECK_EVERY: Duration = Duration::from_secs(15);
pub const SKIP_IF_SUCCESS_BEFORE: Duration = Duration::from_secs(5);
pub const CHECK_TIMEOUT_DURATION: Duration = Duration::from_secs(5);
lazy_static::lazy_static! {
static ref TEST_DATA: Bytes = Bytes::from(vec![42u8; 2048]);
@@ -104,6 +102,20 @@ pub fn get_drive_walkdir_stall_timeout() -> Duration {
)
}
pub fn get_drive_active_check_interval() -> Duration {
Duration::from_secs(rustfs_utils::get_env_u64(
rustfs_config::ENV_DRIVE_ACTIVE_CHECK_INTERVAL_SECS,
rustfs_config::DEFAULT_DRIVE_ACTIVE_CHECK_INTERVAL_SECS,
))
}
pub fn get_drive_active_check_timeout() -> Duration {
Duration::from_secs(rustfs_utils::get_env_u64(
rustfs_config::ENV_DRIVE_ACTIVE_CHECK_TIMEOUT_SECS,
rustfs_config::DEFAULT_DRIVE_ACTIVE_CHECK_TIMEOUT_SECS,
))
}
/// DiskHealthTracker tracks the health status of a disk.
/// Similar to Go's diskHealthTracker.
#[derive(Debug)]
@@ -474,7 +486,7 @@ impl LocalDiskWrapper {
async fn monitor_disk_writable(disk: Arc<LocalDisk>, health: Arc<DiskHealthTracker>, cancel_token: CancellationToken) {
// TODO: config interval
let mut interval = time::interval(CHECK_EVERY);
let mut interval = time::interval(get_drive_active_check_interval());
loop {
tokio::select! {
@@ -507,7 +519,16 @@ impl LocalDiskWrapper {
let test_obj = format!("health-check-{}", Uuid::new_v4());
if Self::perform_health_check(disk.clone(), &TEST_BUCKET, &test_obj, &TEST_DATA, true, CHECK_TIMEOUT_DURATION).await.is_err()
if Self::perform_health_check(
disk.clone(),
&TEST_BUCKET,
&test_obj,
&TEST_DATA,
true,
get_drive_active_check_timeout(),
)
.await
.is_err()
&& health.mark_failure(&disk.endpoint(), "active_health_check_failed")
{
// Health check failed, disk is considered faulty
@@ -613,7 +634,16 @@ impl LocalDiskWrapper {
}
let test_obj = format!("health-check-{}", Uuid::new_v4());
match Self::perform_health_check(disk.clone(), &TEST_BUCKET, &test_obj, &TEST_DATA, false, CHECK_TIMEOUT_DURATION).await {
match Self::perform_health_check(
disk.clone(),
&TEST_BUCKET,
&test_obj,
&TEST_DATA,
false,
get_drive_active_check_timeout(),
)
.await
{
Ok(_) => {
let state_before = health.runtime_state();
let is_online = health.mark_recovery_success(&disk.endpoint(), "recovery_probe_success");
@@ -1137,6 +1167,40 @@ mod tests {
});
}
#[test]
fn drive_active_check_interval_uses_default_when_unset() {
temp_env::with_var_unset(rustfs_config::ENV_DRIVE_ACTIVE_CHECK_INTERVAL_SECS, || {
assert_eq!(
get_drive_active_check_interval(),
Duration::from_secs(rustfs_config::DEFAULT_DRIVE_ACTIVE_CHECK_INTERVAL_SECS)
);
});
}
#[test]
fn drive_active_check_interval_reads_env_override() {
temp_env::with_var(rustfs_config::ENV_DRIVE_ACTIVE_CHECK_INTERVAL_SECS, Some("3"), || {
assert_eq!(get_drive_active_check_interval(), Duration::from_secs(3));
});
}
#[test]
fn drive_active_check_timeout_uses_default_when_unset() {
temp_env::with_var_unset(rustfs_config::ENV_DRIVE_ACTIVE_CHECK_TIMEOUT_SECS, || {
assert_eq!(
get_drive_active_check_timeout(),
Duration::from_secs(rustfs_config::DEFAULT_DRIVE_ACTIVE_CHECK_TIMEOUT_SECS)
);
});
}
#[test]
fn drive_active_check_timeout_reads_env_override() {
temp_env::with_var(rustfs_config::ENV_DRIVE_ACTIVE_CHECK_TIMEOUT_SECS, Some("1"), || {
assert_eq!(get_drive_active_check_timeout(), Duration::from_secs(1));
});
}
#[test]
fn runtime_state_transitions_from_online_to_suspect_then_offline() {
temp_env::with_var(rustfs_config::ENV_DRIVE_SUSPECT_FAILURE_THRESHOLD, Some("2"), || {