fix: sector counter for recovery
This commit is contained in:
parent
f8cf3b1510
commit
650130f9c4
@ -58,7 +58,7 @@ func (s *WindowPoStScheduler) declareRecoveries(ctx context.Context, dlIdx uint6
|
|||||||
|
|
||||||
var batchedRecoveryDecls [][]miner.RecoveryDeclaration
|
var batchedRecoveryDecls [][]miner.RecoveryDeclaration
|
||||||
batchedRecoveryDecls = append(batchedRecoveryDecls, []miner.RecoveryDeclaration{})
|
batchedRecoveryDecls = append(batchedRecoveryDecls, []miner.RecoveryDeclaration{})
|
||||||
totalRecoveries := 0
|
totalSectorsToRecover := uint64(0)
|
||||||
|
|
||||||
for partIdx, partition := range partitions {
|
for partIdx, partition := range partitions {
|
||||||
unrecovered, err := bitfield.SubtractBitField(partition.FaultySectors, partition.RecoveringSectors)
|
unrecovered, err := bitfield.SubtractBitField(partition.FaultySectors, partition.RecoveringSectors)
|
||||||
@ -104,15 +104,17 @@ func (s *WindowPoStScheduler) declareRecoveries(ctx context.Context, dlIdx uint6
|
|||||||
Sectors: recovered,
|
Sectors: recovered,
|
||||||
})
|
})
|
||||||
|
|
||||||
if recoveringSectorLimit > 0 && int64(totalRecoveries) >= recoveringSectorLimit {
|
totalSectorsToRecover += recoveredCount
|
||||||
log.Errorw("reached recovering sector limit, not all sectors will be marked as recovered")
|
|
||||||
|
if recoveringSectorLimit > 0 && int64(totalSectorsToRecover) >= recoveringSectorLimit {
|
||||||
|
log.Errorf("reached recovering sector limit %d, only marking %d sectors for recovery now",
|
||||||
|
recoveringSectorLimit,
|
||||||
|
totalSectorsToRecover)
|
||||||
break
|
break
|
||||||
}
|
}
|
||||||
|
|
||||||
totalRecoveries++
|
|
||||||
}
|
}
|
||||||
|
|
||||||
if totalRecoveries == 0 {
|
if totalSectorsToRecover == 0 {
|
||||||
if faulty != 0 {
|
if faulty != 0 {
|
||||||
log.Warnw("No recoveries to declare", "deadline", dlIdx, "faulty", faulty)
|
log.Warnw("No recoveries to declare", "deadline", dlIdx, "faulty", faulty)
|
||||||
}
|
}
|
||||||
@ -120,6 +122,7 @@ func (s *WindowPoStScheduler) declareRecoveries(ctx context.Context, dlIdx uint6
|
|||||||
return nil, nil, nil
|
return nil, nil, nil
|
||||||
}
|
}
|
||||||
|
|
||||||
|
log.Infof("attempting recovery declarations for %d sectors", totalSectorsToRecover)
|
||||||
var msgs []*types.SignedMessage
|
var msgs []*types.SignedMessage
|
||||||
for _, recovery := range batchedRecoveryDecls {
|
for _, recovery := range batchedRecoveryDecls {
|
||||||
params := &miner.DeclareFaultsRecoveredParams{
|
params := &miner.DeclareFaultsRecoveredParams{
|
||||||
@ -141,7 +144,6 @@ func (s *WindowPoStScheduler) declareRecoveries(ctx context.Context, dlIdx uint6
|
|||||||
if err := s.prepareMessage(ctx, msg, spec); err != nil {
|
if err := s.prepareMessage(ctx, msg, spec); err != nil {
|
||||||
return nil, nil, err
|
return nil, nil, err
|
||||||
}
|
}
|
||||||
|
|
||||||
sm, err := s.api.MpoolPushMessage(ctx, msg, &api.MessageSendSpec{MaxFee: abi.TokenAmount(s.feeCfg.MaxWindowPoStGasFee)})
|
sm, err := s.api.MpoolPushMessage(ctx, msg, &api.MessageSendSpec{MaxFee: abi.TokenAmount(s.feeCfg.MaxWindowPoStGasFee)})
|
||||||
if err != nil {
|
if err != nil {
|
||||||
return nil, nil, xerrors.Errorf("pushing message to mpool: %w", err)
|
return nil, nil, xerrors.Errorf("pushing message to mpool: %w", err)
|
||||||
|
Loading…
Reference in New Issue
Block a user