You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
rocksdb/logging/env_logger.h

166 lines
4.7 KiB

// Copyright (c) 2011-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).
//
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
//
// Logger implementation that uses custom Env object for logging.
#pragma once
#include <time.h>
#include <atomic>
#include <memory>
#include "port/sys_time.h"
#include "file/writable_file_writer.h"
#include "monitoring/iostats_context_imp.h"
#include "rocksdb/env.h"
#include "rocksdb/slice.h"
#include "test_util/sync_point.h"
#include "util/mutexlock.h"
namespace ROCKSDB_NAMESPACE {
class EnvLogger : public Logger {
public:
Introduce a new storage specific Env API (#5761) Summary: The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc. This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO. The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before. This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection. The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761 Differential Revision: D18868376 Pulled By: anand1976 fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
5 years ago
EnvLogger(std::unique_ptr<FSWritableFile>&& writable_file,
const std::string& fname, const EnvOptions& options, Env* env,
InfoLogLevel log_level = InfoLogLevel::ERROR_LEVEL)
: Logger(log_level),
file_(std::move(writable_file), fname, options, env->GetSystemClock()),
last_flush_micros_(0),
env_(env),
flush_pending_(false) {}
~EnvLogger() {
if (!closed_) {
closed_ = true;
CloseHelper().PermitUncheckedError();
}
}
private:
void FlushLocked() {
mutex_.AssertHeld();
if (flush_pending_) {
flush_pending_ = false;
file_.Flush().PermitUncheckedError();
}
last_flush_micros_ = env_->NowMicros();
}
void Flush() override {
TEST_SYNC_POINT("EnvLogger::Flush:Begin1");
TEST_SYNC_POINT("EnvLogger::Flush:Begin2");
MutexLock l(&mutex_);
FlushLocked();
}
Status CloseImpl() override { return CloseHelper(); }
Status CloseHelper() {
mutex_.Lock();
const auto close_status = file_.Close();
mutex_.Unlock();
if (close_status.ok()) {
return close_status;
}
return Status::IOError("Close of log file failed with error:" +
(close_status.getState()
? std::string(close_status.getState())
: std::string()));
}
using Logger::Logv;
void Logv(const char* format, va_list ap) override {
IOSTATS_TIMER_GUARD(logger_nanos);
const uint64_t thread_id = env_->GetThreadID();
// We try twice: the first time with a fixed-size stack allocated buffer,
// and the second time with a much larger dynamically allocated buffer.
char buffer[500];
for (int iter = 0; iter < 2; iter++) {
char* base;
int bufsize;
if (iter == 0) {
bufsize = sizeof(buffer);
base = buffer;
} else {
bufsize = 65536;
base = new char[bufsize];
}
char* p = base;
char* limit = base + bufsize;
struct timeval now_tv;
gettimeofday(&now_tv, nullptr);
const time_t seconds = now_tv.tv_sec;
struct tm t;
localtime_r(&seconds, &t);
p += snprintf(p, limit - p, "%04d/%02d/%02d-%02d:%02d:%02d.%06d %llx ",
t.tm_year + 1900, t.tm_mon + 1, t.tm_mday, t.tm_hour,
t.tm_min, t.tm_sec, static_cast<int>(now_tv.tv_usec),
static_cast<long long unsigned int>(thread_id));
// Print the message
if (p < limit) {
va_list backup_ap;
va_copy(backup_ap, ap);
p += vsnprintf(p, limit - p, format, backup_ap);
va_end(backup_ap);
}
// Truncate to available space if necessary
if (p >= limit) {
if (iter == 0) {
continue; // Try again with larger buffer
} else {
p = limit - 1;
}
}
// Add newline if necessary
if (p == base || p[-1] != '\n') {
*p++ = '\n';
}
assert(p <= limit);
mutex_.Lock();
// We will ignore any error returned by Append().
file_.Append(Slice(base, p - base)).PermitUncheckedError();
flush_pending_ = true;
const uint64_t now_micros = env_->NowMicros();
if (now_micros - last_flush_micros_ >= flush_every_seconds_ * 1000000) {
FlushLocked();
}
mutex_.Unlock();
if (base != buffer) {
delete[] base;
}
break;
}
}
size_t GetLogFileSize() const override {
MutexLock l(&mutex_);
return file_.GetFileSize();
}
private:
WritableFileWriter file_;
mutable port::Mutex mutex_; // Mutex to protect the shared variables below.
const static uint64_t flush_every_seconds_ = 5;
std::atomic_uint_fast64_t last_flush_micros_;
Env* env_;
std::atomic<bool> flush_pending_;
};
} // namespace ROCKSDB_NAMESPACE