forked from facebook/rocksdb
-
Notifications
You must be signed in to change notification settings - Fork 0
/
error_handler.h
150 lines (120 loc) · 5.09 KB
/
error_handler.h
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
// Copyright (c) 2018-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
#pragma once
#include <sstream>
#include "monitoring/instrumented_mutex.h"
#include "options/db_options.h"
#include "rocksdb/io_status.h"
#include "rocksdb/listener.h"
#include "rocksdb/status.h"
#include "util/autovector.h"
namespace ROCKSDB_NAMESPACE {
class DBImpl;
// This structure is used to store the DB recovery context. The context is
// the information that related to the recover actions. For example, it contains
// FlushReason, which tells the flush job why this flush is called.
struct DBRecoverContext {
FlushReason flush_reason;
bool flush_after_recovery;
DBRecoverContext()
: flush_reason(FlushReason::kErrorRecovery),
flush_after_recovery(false) {}
DBRecoverContext(FlushReason reason)
: flush_reason(reason), flush_after_recovery(false) {}
};
class ErrorHandler {
public:
ErrorHandler(DBImpl* db, const ImmutableDBOptions& db_options,
InstrumentedMutex* db_mutex)
: db_(db),
db_options_(db_options),
cv_(db_mutex),
end_recovery_(false),
recovery_thread_(nullptr),
db_mutex_(db_mutex),
auto_recovery_(false),
recovery_in_prog_(false),
soft_error_no_bg_work_(false),
is_db_stopped_(false),
bg_error_stats_(db_options.statistics) {
// Clear the checked flag for uninitialized errors
bg_error_.PermitUncheckedError();
recovery_error_.PermitUncheckedError();
}
void EnableAutoRecovery() { auto_recovery_ = true; }
Status::Severity GetErrorSeverity(BackgroundErrorReason reason,
Status::Code code, Status::SubCode subcode);
const Status& SetBGError(const Status& bg_err, BackgroundErrorReason reason);
Status GetBGError() const { return bg_error_; }
Status GetRecoveryError() const { return recovery_error_; }
Status ClearBGError();
bool IsDBStopped() { return is_db_stopped_.load(std::memory_order_acquire); }
bool IsBGWorkStopped() {
assert(db_mutex_);
db_mutex_->AssertHeld();
return !bg_error_.ok() &&
(bg_error_.severity() >= Status::Severity::kHardError ||
!auto_recovery_ || soft_error_no_bg_work_);
}
bool IsSoftErrorNoBGWork() { return soft_error_no_bg_work_; }
bool IsRecoveryInProgress() { return recovery_in_prog_; }
Status RecoverFromBGError(bool is_manual = false);
void CancelErrorRecovery();
void EndAutoRecovery();
void AddFilesToQuarantine(
autovector<const autovector<uint64_t>*> files_to_quarantine);
const autovector<uint64_t>& GetFilesToQuarantine() const {
db_mutex_->AssertHeld();
return files_to_quarantine_;
}
void ClearFilesToQuarantine();
private:
void RecordStats(
const std::vector<Tickers>& ticker_types,
const std::vector<std::tuple<Histograms, uint64_t>>& int_histograms);
DBImpl* db_;
const ImmutableDBOptions& db_options_;
Status bg_error_;
// A separate Status variable used to record any errors during the
// recovery process from hard errors
IOStatus recovery_error_;
// The condition variable used with db_mutex during auto resume for time
// wait.
InstrumentedCondVar cv_;
bool end_recovery_;
std::unique_ptr<port::Thread> recovery_thread_;
InstrumentedMutex* db_mutex_;
// A flag indicating whether automatic recovery from errors is enabled. Auto
// recovery applies for delegating to SstFileManager to handle no space type
// of errors. This flag doesn't control the auto resume behavior to recover
// from retryable IO errors.
bool auto_recovery_;
bool recovery_in_prog_;
// A flag to indicate that for the soft error, we should not allow any
// background work except the work is from recovery.
bool soft_error_no_bg_work_;
// Used to store the context for recover, such as flush reason.
DBRecoverContext recover_context_;
std::atomic<bool> is_db_stopped_;
// The pointer of DB statistics.
std::shared_ptr<Statistics> bg_error_stats_;
// During recovery from manifest IO errors, files whose VersionEdits entries
// could be in an ambiguous state are quarantined and file deletion refrain
// from deleting them. Successful recovery will clear this vector. Files are
// added to this vector while DB mutex was locked, this data structure is
// unsorted.
autovector<uint64_t> files_to_quarantine_;
const Status& HandleKnownErrors(const Status& bg_err,
BackgroundErrorReason reason);
Status OverrideNoSpaceError(const Status& bg_error, bool* auto_recovery);
void RecoverFromNoSpace();
const Status& StartRecoverFromRetryableBGIOError(const IOStatus& io_error);
void RecoverFromRetryableBGIOError();
// First, if it is in recovery and the recovery_error is ok. Set the
// recovery_error_ to bg_err. Second, if the severity is higher than the
// current bg_error_, overwrite it.
void CheckAndSetRecoveryAndBGError(const Status& bg_err);
};
} // namespace ROCKSDB_NAMESPACE