mirror of
https://github.com/fawney19/Aether.git
synced 2026-09-12 22:20:19 +08:00
fix(gateway): complete worker registration cleanup
This commit is contained in:
elky
+22
@@ -0,0 +1,22 @@
|
||||
-- Worker supervisors are registered as one logical row per task. Older binaries
|
||||
-- included the ephemeral gateway instance in the row id, leaving a permanently
|
||||
-- running row after every restart. Remove only those system-generated boot rows;
|
||||
-- current workers recreate the stable logical rows after migrations complete.
|
||||
-- The metadata predicate also replaces task-only rows written by early builds of
|
||||
-- this fix that still claimed an instance owner. Delete children explicitly so
|
||||
-- cleanup remains complete after imports performed with FK checks disabled.
|
||||
DELETE FROM background_task_events
|
||||
WHERE run_id IN (
|
||||
SELECT id
|
||||
FROM background_task_runs
|
||||
WHERE id LIKE 'boot:%'
|
||||
AND owner_instance IS NOT NULL
|
||||
AND created_by = 'system'
|
||||
AND progress_message = 'worker booted'
|
||||
);
|
||||
|
||||
DELETE FROM background_task_runs
|
||||
WHERE id LIKE 'boot:%'
|
||||
AND owner_instance IS NOT NULL
|
||||
AND created_by = 'system'
|
||||
AND progress_message = 'worker booted';
|
||||
+22
@@ -0,0 +1,22 @@
|
||||
-- Worker supervisors are registered as one logical row per task. Older binaries
|
||||
-- included the ephemeral gateway instance in the row id, leaving a permanently
|
||||
-- running row after every restart. Remove only those system-generated boot rows;
|
||||
-- current workers recreate the stable logical rows after migrations complete.
|
||||
-- The metadata predicate also replaces task-only rows written by early builds of
|
||||
-- this fix that still claimed an instance owner. Delete children explicitly so
|
||||
-- cleanup remains complete after imports performed with FK checks disabled.
|
||||
DELETE FROM background_task_events
|
||||
WHERE run_id IN (
|
||||
SELECT id
|
||||
FROM background_task_runs
|
||||
WHERE id LIKE 'boot:%'
|
||||
AND owner_instance IS NOT NULL
|
||||
AND created_by = 'system'
|
||||
AND progress_message = 'worker booted'
|
||||
);
|
||||
|
||||
DELETE FROM background_task_runs
|
||||
WHERE id LIKE 'boot:%'
|
||||
AND owner_instance IS NOT NULL
|
||||
AND created_by = 'system'
|
||||
AND progress_message = 'worker booted';
|
||||
+22
@@ -0,0 +1,22 @@
|
||||
-- Worker supervisors are registered as one logical row per task. Older binaries
|
||||
-- included the ephemeral gateway instance in the row id, leaving a permanently
|
||||
-- running row after every restart. Remove only those system-generated boot rows;
|
||||
-- current workers recreate the stable logical rows after migrations complete.
|
||||
-- The metadata predicate also replaces task-only rows written by early builds of
|
||||
-- this fix that still claimed an instance owner. Delete children explicitly so
|
||||
-- cleanup remains complete after imports performed with FK checks disabled.
|
||||
DELETE FROM background_task_events
|
||||
WHERE run_id IN (
|
||||
SELECT id
|
||||
FROM background_task_runs
|
||||
WHERE id LIKE 'boot:%'
|
||||
AND owner_instance IS NOT NULL
|
||||
AND created_by = 'system'
|
||||
AND progress_message = 'worker booted'
|
||||
);
|
||||
|
||||
DELETE FROM background_task_runs
|
||||
WHERE id LIKE 'boot:%'
|
||||
AND owner_instance IS NOT NULL
|
||||
AND created_by = 'system'
|
||||
AND progress_message = 'worker booted';
|
||||
@@ -538,6 +538,146 @@ WHERE request_id = ?
|
||||
assert_eq!(remaining, 0);
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn worker_boot_cleanup_removes_only_legacy_instance_rows_and_events() {
|
||||
const CLEANUP_VERSION: i64 = 20260731000000;
|
||||
const HASHED_LEGACY_ID: &str =
|
||||
"boot:maintenance.request.candidate.cleanup:~0123456789abcdef0123";
|
||||
const OVERLONG_LEGACY_ID: &str =
|
||||
"boot:maintenance.proxy.node.metrics.cleanup:gateway-instance-with-an-overlong-id";
|
||||
|
||||
assert_eq!(HASHED_LEGACY_ID.len(), 64);
|
||||
assert!(OVERLONG_LEGACY_ID.len() > 64);
|
||||
|
||||
let pool = sqlx::sqlite::SqlitePoolOptions::new()
|
||||
.max_connections(1)
|
||||
.connect("sqlite::memory:")
|
||||
.await
|
||||
.expect("in-memory sqlite pool");
|
||||
|
||||
for migration in MIGRATOR
|
||||
.iter()
|
||||
.filter(|migration| migration.version < CLEANUP_VERSION)
|
||||
{
|
||||
sqlx::raw_sql(migration.sql.as_ref())
|
||||
.execute(&pool)
|
||||
.await
|
||||
.unwrap_or_else(|err| panic!("migration {} should run: {err}", migration.version));
|
||||
}
|
||||
|
||||
sqlx::query(
|
||||
r#"
|
||||
INSERT INTO background_task_runs (
|
||||
id, task_key, kind, "trigger", status, owner_instance,
|
||||
progress_message, created_by, created_at_unix_secs, updated_at_unix_secs
|
||||
) VALUES
|
||||
('boot:usage.queue.worker:gateway-a', 'usage.queue.worker', 'daemon', 'daemon',
|
||||
'running', 'gateway-a', 'worker booted', 'system', 1, 1),
|
||||
('boot:usage.queue.worker:gateway-b', 'usage.queue.worker', 'daemon', 'daemon',
|
||||
'running', 'gateway-b', 'worker booted', 'system', 2, 2),
|
||||
('boot:maintenance.request.candidate.cleanup:~0123456789abcdef0123',
|
||||
'maintenance.request.candidate.cleanup', 'scheduled', 'interval',
|
||||
'running', 'gateway-hash', 'worker booted', 'system', 3, 3),
|
||||
('boot:maintenance.proxy.node.metrics.cleanup:gateway-instance-with-an-overlong-id',
|
||||
'maintenance.proxy.node.metrics.cleanup', 'scheduled', 'interval',
|
||||
'running', 'gateway-overlong', 'worker booted', 'system', 4, 4),
|
||||
('boot:model.fetch.worker', 'model.fetch.worker', 'scheduled', 'interval',
|
||||
'running', 'gateway-early-fix', 'worker booted', 'system', 5, 5),
|
||||
('boot:usage.queue.worker', 'usage.queue.worker', 'daemon', 'daemon',
|
||||
'running', NULL, 'worker registered', 'system', 6, 6),
|
||||
('boot:ownerless-worker', 'ownerless.worker', 'daemon', 'daemon',
|
||||
'running', NULL, 'worker booted', 'system', 7, 7),
|
||||
('boot:custom-progress', 'custom.progress', 'daemon', 'daemon',
|
||||
'running', 'gateway-custom', 'worker healthy', 'system', 8, 8),
|
||||
('boot:user-request', 'user.request', 'on_demand', 'manual',
|
||||
'running', 'gateway-user', 'worker booted', 'admin', 9, 9);
|
||||
|
||||
INSERT INTO background_task_events (
|
||||
id, run_id, event_type, message, created_at_unix_secs
|
||||
) VALUES
|
||||
('legacy-event-a', 'boot:usage.queue.worker:gateway-a', 'worker_boot', 'legacy', 1),
|
||||
('legacy-event-b', 'boot:usage.queue.worker:gateway-b', 'worker_boot', 'legacy', 2),
|
||||
('legacy-event-hash', 'boot:maintenance.request.candidate.cleanup:~0123456789abcdef0123',
|
||||
'worker_boot', 'legacy hash', 3),
|
||||
('legacy-event-overlong',
|
||||
'boot:maintenance.proxy.node.metrics.cleanup:gateway-instance-with-an-overlong-id',
|
||||
'worker_boot', 'legacy overlong', 4),
|
||||
('early-fix-event', 'boot:model.fetch.worker', 'worker_boot', 'early fix', 5),
|
||||
('logical-event', 'boot:usage.queue.worker', 'worker_boot', 'logical', 6),
|
||||
('ownerless-event', 'boot:ownerless-worker', 'worker_boot', 'ownerless', 7),
|
||||
('custom-progress-event', 'boot:custom-progress', 'worker_boot', 'custom progress', 8),
|
||||
('manual-event', 'boot:user-request', 'manual', 'manual', 9);
|
||||
"#,
|
||||
)
|
||||
.execute(&pool)
|
||||
.await
|
||||
.expect("worker boot cleanup fixtures should insert");
|
||||
|
||||
let migration = MIGRATOR
|
||||
.iter()
|
||||
.find(|migration| migration.version == CLEANUP_VERSION)
|
||||
.expect("worker boot cleanup migration should be embedded");
|
||||
sqlx::raw_sql(migration.sql.as_ref())
|
||||
.execute(&pool)
|
||||
.await
|
||||
.expect("worker boot cleanup migration should run");
|
||||
sqlx::raw_sql(migration.sql.as_ref())
|
||||
.execute(&pool)
|
||||
.await
|
||||
.expect("worker boot cleanup migration should be idempotent");
|
||||
|
||||
let remaining_runs = sqlx::query_as::<_, (String, Option<String>, Option<String>)>(
|
||||
r#"
|
||||
SELECT id, owner_instance, created_by
|
||||
FROM background_task_runs
|
||||
ORDER BY id
|
||||
"#,
|
||||
)
|
||||
.fetch_all(&pool)
|
||||
.await
|
||||
.expect("remaining worker task rows should load");
|
||||
assert_eq!(
|
||||
remaining_runs,
|
||||
vec![
|
||||
(
|
||||
"boot:custom-progress".to_string(),
|
||||
Some("gateway-custom".to_string()),
|
||||
Some("system".to_string()),
|
||||
),
|
||||
(
|
||||
"boot:ownerless-worker".to_string(),
|
||||
None,
|
||||
Some("system".to_string()),
|
||||
),
|
||||
(
|
||||
"boot:usage.queue.worker".to_string(),
|
||||
None,
|
||||
Some("system".to_string()),
|
||||
),
|
||||
(
|
||||
"boot:user-request".to_string(),
|
||||
Some("gateway-user".to_string()),
|
||||
Some("admin".to_string()),
|
||||
),
|
||||
]
|
||||
);
|
||||
|
||||
let remaining_events =
|
||||
sqlx::query_scalar::<_, String>("SELECT id FROM background_task_events ORDER BY id")
|
||||
.fetch_all(&pool)
|
||||
.await
|
||||
.expect("remaining worker task events should load");
|
||||
assert_eq!(
|
||||
remaining_events,
|
||||
vec![
|
||||
"custom-progress-event".to_string(),
|
||||
"logical-event".to_string(),
|
||||
"manual-event".to_string(),
|
||||
"ownerless-event".to_string(),
|
||||
]
|
||||
);
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn pending_and_startup_preparation_reject_dirty_migration_state() {
|
||||
let pool = sqlx::sqlite::SqlitePoolOptions::new()
|
||||
|
||||
Reference in New Issue
Block a user