astelm/minio
1
0
Fork 0
mirror of https://github.com/pgsty/minio.git synced 2026-11-11 01:55:30 +03:00
Code Issues Packages Projects Releases Wiki Activity

fix(heal): heal a version whose expiry drive healing enqueues

Browse Source 
Drive healing evaluates lifecycle on the versions of one pool. For a
version it found due it enqueued the expiry and skipped the heal. The
expiry is decided again under the object lock, with the versions in all
pools and the Object Lock state it reads then, and can be refused: a
newer version in another pool, an unreadable history, a changed rule,
or a delete-all the exact-key purge refuses. Such a version was never
healed onto the replaced drive and kept reduced redundancy for as long
as the refusal lasted. The scanner heals only versions without a due
action, so it did not repair it either.

Keep the enqueue and heal the version. A version that the expiry then
removes costs one heal. Heal and expiry exclude each other through the
object lock in the writers' lock domain, which the single-pool lock
domain change provides; a version removed first reads as not found and
is skipped, and heal rebuilds only from read-quorum metadata, so it does
not bring back a version deleted on a write quorum.

Tests: a version refused across pools is healed onto the replaced drive
and survives the expiry; an unversioned object due is healed and then
removed from every drive; a versioned latest version due is healed and
hidden by a marker; a stale minority is not recreated. The first three
fail without the change.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Feng Ruohang <rh@vonng.com>
This commit is contained in:
Feng Ruohang
2026-11-06 18:22:06 +08:00
parent f4b6c04c9f
commit 928c827508
2 changed files with 197 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
cmd/global-heal.go
+11 -4
View File
@@ -371,8 +371,14 @@ func (er *erasureObjects) healErasureSet(ctx context.Context, buckets []string,
applyExpiryOnTransitionedObject(ctx, newObjectLayerFn(), objInfo, evt, lcEventSrc_Heal)
return false
case evt.Action.Delete():
// Enqueue the expiry, but heal the version: the expiry is
// decided again under the object lock, with the versions in
// all pools and the protection it reads then, and can be
// refused. A skipped version would stay without this drive
// for as long as it is refused. One that is deleted later
// costs one heal.
globalExpiryState.enqueueByDays(objInfo, evt, lcEventSrc_Heal)
return true
return false
default:
return false
}
@@ -450,9 +456,10 @@ func (er *erasureObjects) healErasureSet(ctx context.Context, buckets []string,
var versionNotFound int
for _, version := range fivs.Versions {
// Ignore healing a version if:
// - It is uploaded after the drive healing is started
// - An object that is already expired by ILM rule.
// Ignore healing a version that is uploaded after the drive
// healing started. filterLifecycle only schedules the ILM
// action due for the version; it reports false for every
// version, also one it expires.
if !started.IsZero() && version.ModTime.After(started) || filterLifecycle(bucket, version.Name, version) {
versionNotFound++
if !send(healEntrySkipped(uint64(version.Size))) {