mirror of
https://github.com/minio/minio.git
synced 2024-12-26 23:25:54 -05:00
9a012a53ef
This PR fixes a bug that perhaps has been long introduced, with no visible workarounds. In any deployment, if an entire erasure set is deleted, there is no way the cluster recovers.
105 lines
2.6 KiB
Go
105 lines
2.6 KiB
Go
// Copyright (c) 2015-2021 MinIO, Inc.
|
|
//
|
|
// This file is part of MinIO Object Storage stack
|
|
//
|
|
// This program is free software: you can redistribute it and/or modify
|
|
// it under the terms of the GNU Affero General Public License as published by
|
|
// the Free Software Foundation, either version 3 of the License, or
|
|
// (at your option) any later version.
|
|
//
|
|
// This program is distributed in the hope that it will be useful
|
|
// but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
// GNU Affero General Public License for more details.
|
|
//
|
|
// You should have received a copy of the GNU Affero General Public License
|
|
// along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
package cmd
|
|
|
|
import (
|
|
"context"
|
|
"time"
|
|
|
|
"github.com/minio/madmin-go/v3"
|
|
)
|
|
|
|
const (
|
|
mrfOpsQueueSize = 100000
|
|
)
|
|
|
|
// partialOperation is a successful upload/delete of an object
|
|
// but not written in all disks (having quorum)
|
|
type partialOperation struct {
|
|
bucket string
|
|
object string
|
|
versionID string
|
|
allVersions bool
|
|
setIndex, poolIndex int
|
|
queued time.Time
|
|
scanMode madmin.HealScanMode
|
|
}
|
|
|
|
// mrfState sncapsulates all the information
|
|
// related to the global background MRF.
|
|
type mrfState struct {
|
|
opCh chan partialOperation
|
|
}
|
|
|
|
// Add a partial S3 operation (put/delete) when one or more disks are offline.
|
|
func (m *mrfState) addPartialOp(op partialOperation) {
|
|
if m == nil {
|
|
return
|
|
}
|
|
|
|
select {
|
|
case m.opCh <- op:
|
|
default:
|
|
}
|
|
}
|
|
|
|
var healSleeper = newDynamicSleeper(5, time.Second, false)
|
|
|
|
// healRoutine listens to new disks reconnection events and
|
|
// issues healing requests for queued objects belonging to the
|
|
// corresponding erasure set
|
|
func (m *mrfState) healRoutine(z *erasureServerPools) {
|
|
for {
|
|
select {
|
|
case <-GlobalContext.Done():
|
|
return
|
|
case u, ok := <-m.opCh:
|
|
if !ok {
|
|
return
|
|
}
|
|
|
|
now := time.Now()
|
|
if now.Sub(u.queued) < time.Second {
|
|
// let recently failed networks to reconnect
|
|
// making MRF wait for 1s before retrying,
|
|
// i.e 4 reconnect attempts.
|
|
time.Sleep(time.Second)
|
|
}
|
|
|
|
// wait on timer per heal
|
|
wait := healSleeper.Timer(context.Background())
|
|
|
|
scan := madmin.HealNormalScan
|
|
if u.scanMode != 0 {
|
|
scan = u.scanMode
|
|
}
|
|
if u.object == "" {
|
|
healBucket(u.bucket, scan)
|
|
} else {
|
|
if u.allVersions {
|
|
z.serverPools[u.poolIndex].sets[u.setIndex].listAndHeal(u.bucket, u.object, u.scanMode, healObjectVersionsDisparity)
|
|
} else {
|
|
healObject(u.bucket, u.object, u.versionID, scan)
|
|
}
|
|
}
|
|
|
|
wait()
|
|
}
|
|
}
|
|
}
|