HDFS-9144. Refactoring libhdfs++ into stateful/ephemeral objects. Contributed by Bob Hansen.

This commit is contained in:
James 2015-12-03 07:30:22 -05:00 committed by James Clampffer
parent 5ba8e8ab7b
commit d7ecf396c9
38 changed files with 1562 additions and 1284 deletions

View File

@ -24,6 +24,7 @@
#include <functional>
#include <memory>
#include <set>
#include <iostream>
namespace hdfs {
@ -68,10 +69,10 @@ class NodeExclusionRule {
};
/**
* Applications opens an InputStream to read files in HDFS.
* Applications opens a FileHandle to read files in HDFS.
**/
class InputStream {
public:
class FileHandle {
public:
/**
* Read data from a specific position. The current implementation
* stops at the block boundary.
@ -83,10 +84,14 @@ class InputStream {
* The handler returns the datanode that serves the block and the number of
* bytes has read.
**/
virtual void PositionRead(
void *buf, size_t nbyte, uint64_t offset,
const std::function<void(const Status &, const std::string &, size_t)> &
handler) = 0;
virtual void
PositionRead(void *buf, size_t nbyte, uint64_t offset,
const std::function<void(const Status &, size_t)> &handler) = 0;
virtual Status PositionRead(void *buf, size_t *nbyte, off_t offset) = 0;
virtual Status Read(void *buf, size_t *nbyte) = 0;
virtual Status Seek(off_t *offset, std::ios_base::seekdir whence) = 0;
/**
* Determine if a datanode should be excluded from future operations
* based on the return Status.
@ -97,7 +102,7 @@ class InputStream {
**/
static bool ShouldExclude(const Status &status);
virtual ~InputStream();
virtual ~FileHandle();
};
/**
@ -114,15 +119,24 @@ class FileSystem {
IoService *io_service, const Options &options, const std::string &server,
const std::string &service,
const std::function<void(const Status &, FileSystem *)> &handler);
/* Synchronous call of New*/
static FileSystem *
New(IoService *io_service, const Options &options, const std::string &server,
const std::string &service);
/**
* Open a file on HDFS. The call issues an RPC to the NameNode to
* gather the locations of all blocks in the file and to return a
* new instance of the @ref InputStream object.
**/
virtual void Open(
const std::string &path,
const std::function<void(const Status &, InputStream *)> &handler) = 0;
virtual ~FileSystem();
virtual void
Open(const std::string &path,
const std::function<void(const Status &, FileHandle *)> &handler) = 0;
virtual Status Open(const std::string &path, FileHandle **handle) = 0;
virtual ~FileSystem() {};
};
}

View File

@ -30,6 +30,17 @@ struct Options {
**/
int rpc_timeout;
/**
* Maximum number of retries for RPC operations
**/
const static int NO_RPC_RETRY = -1;
int max_rpc_retries;
/**
* Number of ms to wait between retry of RPC operations
**/
int rpc_retry_delay_ms;
/**
* Exclusion time for failed datanodes in milliseconds.
* Default: 60000

View File

@ -21,4 +21,5 @@ add_subdirectory(fs)
add_subdirectory(reader)
add_subdirectory(rpc)
add_subdirectory(proto)
add_subdirectory(connection)
add_subdirectory(bindings)

View File

@ -16,5 +16,5 @@
# under the License.
add_library(bindings_c hdfs.cc hdfs_cpp.cc)
add_library(bindings_c hdfs.cc)
add_dependencies(bindings_c fs rpc reader proto common fs rpc reader proto common)

View File

@ -16,8 +16,10 @@
* limitations under the License.
*/
#include "hdfs_cpp.h"
#include "fs/filesystem.h"
#include <hdfs/hdfs.h>
#include <string>
#include <cstring>
#include <iostream>
@ -25,15 +27,15 @@ using namespace hdfs;
/* Seperate the handles used by the C api from the C++ API*/
struct hdfs_internal {
hdfs_internal(HadoopFileSystem *p) : filesystem_(p) {}
hdfs_internal(std::unique_ptr<HadoopFileSystem> p)
hdfs_internal(FileSystem *p) : filesystem_(p) {}
hdfs_internal(std::unique_ptr<FileSystem> p)
: filesystem_(std::move(p)) {}
virtual ~hdfs_internal(){};
HadoopFileSystem *get_impl() { return filesystem_.get(); }
const HadoopFileSystem *get_impl() const { return filesystem_.get(); }
FileSystem *get_impl() { return filesystem_.get(); }
const FileSystem *get_impl() const { return filesystem_.get(); }
private:
std::unique_ptr<HadoopFileSystem> filesystem_;
std::unique_ptr<FileSystem> filesystem_;
};
struct hdfsFile_internal {
@ -102,17 +104,23 @@ bool CheckSystemAndHandle(hdfsFS fs, hdfsFile file) {
int hdfsFileIsOpenForRead(hdfsFile file) {
/* files can only be open for reads at the moment, do a quick check */
if (file) {
return file->get_impl()->IsOpenForRead();
return true; // Update implementation when we get file writing
}
return false;
}
hdfsFS hdfsConnect(const char *nn, tPort port) {
HadoopFileSystem *fs = new HadoopFileSystem();
Status stat = fs->Connect(nn, port);
if (!stat.ok()) {
std::string port_as_string = std::to_string(port);
IoService * io_service = IoService::New();
FileSystem *fs = FileSystem::New(io_service, Options(), nn, port_as_string);
if (!fs) {
ReportError(ENODEV, "Unable to connect to NameNode.");
delete fs;
// FileSystem's ctor might take ownership of the io_service; if it does,
// it will null out the pointer
if (io_service)
delete io_service;
return nullptr;
}
return new hdfs_internal(fs);
@ -139,7 +147,7 @@ hdfsFile hdfsOpenFile(hdfsFS fs, const char *path, int flags, int bufferSize,
return nullptr;
}
FileHandle *f = nullptr;
Status stat = fs->get_impl()->OpenFileForRead(path, &f);
Status stat = fs->get_impl()->Open(path, &f);
if (!stat.ok()) {
return nullptr;
}
@ -150,7 +158,6 @@ int hdfsCloseFile(hdfsFS fs, hdfsFile file) {
if (!CheckSystemAndHandle(fs, file)) {
return -1;
}
delete file;
return 0;
}
@ -162,8 +169,8 @@ tSize hdfsPread(hdfsFS fs, hdfsFile file, tOffset position, void *buffer,
}
size_t len = length;
Status stat = file->get_impl()->Pread(buffer, &len, position);
if (!stat.ok()) {
Status stat = file->get_impl()->PositionRead(buffer, &len, position);
if(!stat.ok()) {
return Error(stat);
}
return (tSize)len;

View File

@ -1,216 +0,0 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "hdfs_cpp.h"
#include <cstdint>
#include <cerrno>
#include <string>
#include <future>
#include <memory>
#include <thread>
#include <vector>
#include <set>
#include <tuple>
#include <hdfs/hdfs.h>
#include "libhdfspp/hdfs.h"
#include "libhdfspp/status.h"
#include "fs/filesystem.h"
#include "common/hdfs_public_api.h"
namespace hdfs {
FileHandle::FileHandle(InputStream *is) : input_stream_(is), offset_(0){}
Status FileHandle::Pread(void *buf, size_t *nbyte, off_t offset) {
auto callstate = std::make_shared<std::promise<std::tuple<Status, std::string, size_t>>>();
std::future<std::tuple<Status, std::string, size_t>> future(callstate->get_future());
/* wrap async call with promise/future to make it blocking */
auto callback = [callstate](
const Status &s, const std::string &dn, size_t bytes) {
callstate->set_value(std::make_tuple(s, dn, bytes));
};
input_stream_->PositionRead(buf, *nbyte, offset, callback);
/* wait for async to finish */
auto returnstate = future.get();
auto stat = std::get<0>(returnstate);
if (!stat.ok()) {
/* determine if DN gets marked bad */
if (InputStream::ShouldExclude(stat)) {
InputStreamImpl *impl =
static_cast<InputStreamImpl *>(input_stream_.get());
impl->bad_node_tracker_->AddBadNode(std::get<1>(returnstate));
}
return stat;
}
*nbyte = std::get<2>(returnstate);
return Status::OK();
}
Status FileHandle::Read(void *buf, size_t *nbyte) {
Status stat = Pread(buf, nbyte, offset_);
if (!stat.ok()) {
return stat;
}
offset_ += *nbyte;
return Status::OK();
}
Status FileHandle::Seek(off_t *offset, std::ios_base::seekdir whence) {
off_t new_offset = -1;
switch (whence) {
case std::ios_base::beg:
new_offset = *offset;
break;
case std::ios_base::cur:
new_offset = offset_ + *offset;
break;
case std::ios_base::end:
new_offset = static_cast<InputStreamImpl *>(input_stream_.get())
->get_file_length() +
*offset;
break;
default:
/* unsupported */
return Status::InvalidArgument("Invalid Seek whence argument");
}
if (!CheckSeekBounds(new_offset)) {
return Status::InvalidArgument("Seek offset out of bounds");
}
offset_ = new_offset;
*offset = offset_;
return Status::OK();
}
/* return false if seek will be out of bounds */
bool FileHandle::CheckSeekBounds(ssize_t desired_position) {
ssize_t file_length =
static_cast<InputStreamImpl *>(input_stream_.get())->get_file_length();
if (desired_position < 0 || desired_position >= file_length) {
return false;
}
return true;
}
bool FileHandle::IsOpenForRead() {
/* for now just check if InputStream exists */
if (!input_stream_) {
return false;
}
return true;
}
HadoopFileSystem::~HadoopFileSystem() {
/**
* Note: IoService must be stopped before getting rid of worker threads.
* Once worker threads are joined and deleted the service can be deleted.
**/
file_system_.reset(nullptr);
service_->Stop();
worker_threads_.clear();
service_.reset(nullptr);
}
Status HadoopFileSystem::Connect(const char *nn, tPort port,
unsigned int threads) {
/* IoService::New can return nullptr */
if (!service_) {
return Status::Error("Null IoService");
}
/* spawn background threads for asio delegation */
for (unsigned int i = 0; i < threads; i++) {
AddWorkerThread();
}
/* synchronized */
auto callstate = std::make_shared<std::promise<std::tuple<Status, FileSystem*>>>();
std::future<std::tuple<Status, FileSystem*>> future(callstate->get_future());
auto callback = [callstate](const Status &s, FileSystem *f) {
callstate->set_value(std::make_tuple(s,f));
};
/* dummy options object until this is hooked up to HDFS-9117 */
Options options_object;
FileSystem::New(service_.get(), options_object, nn, std::to_string(port),
callback);
/* block until promise is set */
auto returnstate = future.get();
Status stat = std::get<0>(returnstate);
FileSystem *fs = std::get<1>(returnstate);
/* check and see if it worked */
if (!stat.ok() || !fs) {
service_->Stop();
worker_threads_.clear();
return stat;
}
file_system_ = std::unique_ptr<FileSystem>(fs);
return stat;
}
int HadoopFileSystem::AddWorkerThread() {
auto service_task = [](IoService *service) { service->Run(); };
worker_threads_.push_back(
WorkerPtr(new std::thread(service_task, service_.get())));
return worker_threads_.size();
}
Status HadoopFileSystem::OpenFileForRead(const std::string &path,
FileHandle **handle) {
auto callstate = std::make_shared<std::promise<std::tuple<Status, InputStream*>>>();
std::future<std::tuple<Status, InputStream*>> future(callstate->get_future());
/* wrap async FileSystem::Open with promise to make it a blocking call */
auto h = [callstate](const Status &s, InputStream *is) {
callstate->set_value(std::make_tuple(s, is));
};
file_system_->Open(path, h);
/* block until promise is set */
auto returnstate = future.get();
Status stat = std::get<0>(returnstate);
InputStream *input_stream = std::get<1>(returnstate);
if (!stat.ok()) {
delete input_stream;
return stat;
}
if (!input_stream) {
return stat;
}
*handle = new FileHandle(input_stream);
return stat;
}
}

View File

@ -1,105 +0,0 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef LIBHDFSPP_BINDINGS_HDFSCPP_H
#define LIBHDFSPP_BINDINGS_HDFSCPP_H
#include <cstdint>
#include <thread>
#include <vector>
#include <mutex>
#include <chrono>
#include <iostream>
#include "libhdfspp/hdfs.h"
#include "fs/bad_datanode_tracker.h"
#include <hdfs/hdfs.h>
namespace hdfs {
/**
* Implement a very simple 'it just works' interface in C++
* that provides posix-like file operations + extra stuff for hadoop.
* Then provide very thin C wrappers over each method.
*/
class HadoopFileSystem;
class FileHandle {
public:
virtual ~FileHandle(){};
/**
* Note: The nbyte argument for Read and Pread as well as the
* offset argument for Seek are in/out parameters.
*
* For Read and Pread the value referenced by nbyte should
* be set to the number of bytes to read. Before returning
* the value referenced will be set by the callee to the number
* of bytes that was successfully read.
*
* For Seek the value referenced by offset should be the number
* of bytes to shift from the specified whence position. The
* referenced value will be set to the new offset before returning.
**/
Status Pread(void *buf, size_t *nbyte, off_t offset);
Status Read(void *buf, size_t *nbyte);
Status Seek(off_t *offset, std::ios_base::seekdir whence);
bool IsOpenForRead();
private:
/* handle should only be created by fs */
friend class HadoopFileSystem;
FileHandle(InputStream *is);
bool CheckSeekBounds(ssize_t desired_position);
std::unique_ptr<InputStream> input_stream_;
off_t offset_;
};
class HadoopFileSystem {
public:
HadoopFileSystem() : service_(IoService::New()) {}
virtual ~HadoopFileSystem();
/* attempt to connect to namenode, return false on failure */
Status Connect(const char *nn, tPort port, unsigned int threads = 1);
/* how many worker threads are servicing asio requests */
int WorkerThreadCount() { return worker_threads_.size(); }
/* add a new thread to handle asio requests, return number of threads in pool
*/
int AddWorkerThread();
Status OpenFileForRead(const std::string &path, FileHandle **handle);
private:
std::unique_ptr<IoService> service_;
/* std::thread needs to join before deletion */
struct WorkerDeleter {
void operator()(std::thread *t) {
t->join();
delete t;
}
};
typedef std::unique_ptr<std::thread, WorkerDeleter> WorkerPtr;
std::vector<WorkerPtr> worker_threads_;
std::unique_ptr<FileSystem> file_system_;
};
}
#endif

View File

@ -1 +1,18 @@
add_library(common base64.cc status.cc sasl_digest_md5.cc hdfs_public_api.cc options.cc configuration.cc)
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
add_library(common base64.cc status.cc sasl_digest_md5.cc hdfs_public_api.cc options.cc configuration.cc util.cc)

View File

@ -0,0 +1,49 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef LIB_COMMON_ASYNC_STREAM_H_
#define LIB_COMMON_ASYNC_STREAM_H_
#include <asio.hpp>
namespace hdfs {
typedef asio::mutable_buffers_1 MutableBuffers;
typedef asio::const_buffers_1 ConstBuffers;
/*
* asio-compatible stream implementation.
*
* Lifecycle: should be managed using std::shared_ptr so the object can be
* handed from consumer to consumer
* Threading model: async_read_some and async_write_some are not thread-safe.
*/
class AsyncStream {
public:
virtual void async_read_some(const MutableBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) = 0;
virtual void async_write_some(const ConstBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) = 0;
};
}
#endif

View File

@ -29,7 +29,9 @@
#include <asio/ip/tcp.hpp>
namespace hdfs {
namespace continuation {
namespace asio_continuation {
using namespace continuation;
template <class Stream, class MutableBufferSequence>
class ReadContinuation : public Continuation {

View File

@ -33,7 +33,7 @@ namespace continuation {
template <class Stream, size_t MaxMessageSize = 512>
struct ReadDelimitedPBMessageContinuation : public Continuation {
ReadDelimitedPBMessageContinuation(Stream *stream,
ReadDelimitedPBMessageContinuation(std::shared_ptr<Stream> stream,
::google::protobuf::MessageLite *msg)
: stream_(stream), msg_(msg) {}
@ -56,8 +56,8 @@ struct ReadDelimitedPBMessageContinuation : public Continuation {
}
next(status);
};
asio::async_read(
*stream_, asio::buffer(buf_),
asio::async_read(*stream_,
asio::buffer(buf_),
std::bind(&ReadDelimitedPBMessageContinuation::CompletionHandler, this,
std::placeholders::_1, std::placeholders::_2),
handler);
@ -82,14 +82,14 @@ private:
return offset ? len + offset - transferred : 1;
}
Stream *stream_;
std::shared_ptr<Stream> stream_;
::google::protobuf::MessageLite *msg_;
std::array<char, MaxMessageSize> buf_;
};
template <class Stream>
struct WriteDelimitedPBMessageContinuation : Continuation {
WriteDelimitedPBMessageContinuation(Stream *stream,
WriteDelimitedPBMessageContinuation(std::shared_ptr<Stream> stream,
const google::protobuf::MessageLite *msg)
: stream_(stream), msg_(msg) {}
@ -101,28 +101,25 @@ struct WriteDelimitedPBMessageContinuation : Continuation {
pbio::CodedOutputStream os(&ss);
os.WriteVarint32(size);
msg_->SerializeToCodedStream(&os);
write_coroutine_ =
std::shared_ptr<Continuation>(Write(stream_, asio::buffer(buf_)));
write_coroutine_->Run([next](const Status &stat) { next(stat); });
asio::async_write(*stream_, asio::buffer(buf_), [next](const asio::error_code &ec, size_t) { next(ToStatus(ec)); } );
}
private:
Stream *stream_;
std::shared_ptr<Stream> stream_;
const google::protobuf::MessageLite *msg_;
std::string buf_;
std::shared_ptr<Continuation> write_coroutine_;
};
template <class Stream, size_t MaxMessageSize = 512>
static inline Continuation *
ReadDelimitedPBMessage(Stream *stream, ::google::protobuf::MessageLite *msg) {
ReadDelimitedPBMessage(std::shared_ptr<Stream> stream, ::google::protobuf::MessageLite *msg) {
return new ReadDelimitedPBMessageContinuation<Stream, MaxMessageSize>(stream,
msg);
}
template <class Stream>
static inline Continuation *
WriteDelimitedPBMessage(Stream *stream, ::google::protobuf::MessageLite *msg) {
WriteDelimitedPBMessage(std::shared_ptr<Stream> stream, ::google::protobuf::MessageLite *msg) {
return new WriteDelimitedPBMessageContinuation<Stream>(stream, msg);
}
}

View File

@ -24,20 +24,4 @@ IoService::~IoService() {}
IoService *IoService::New() { return new IoServiceImpl(); }
bool InputStream::ShouldExclude(const Status &s) {
if (s.ok()) {
return false;
}
switch (s.code()) {
/* client side resource exhaustion */
case Status::kResourceUnavailable:
return false;
case Status::kInvalidArgument:
case Status::kUnimplemented:
case Status::kException:
default:
return true;
}
}
}

View File

@ -20,5 +20,6 @@
namespace hdfs {
Options::Options() : rpc_timeout(30000), host_exclusion_duration(600000) {}
Options::Options() : rpc_timeout(30000), max_rpc_retries(0),
rpc_retry_delay_ms(10000), host_exclusion_duration(600000) {}
}

View File

@ -0,0 +1,35 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "common/util.h"
namespace hdfs {
std::string GetRandomClientName() {
unsigned char buf[6] = {
0,
};
RAND_pseudo_bytes(buf, sizeof(buf));
std::stringstream ss;
ss << "libhdfs++_"
<< Base64Encode(std::string(reinterpret_cast<char *>(buf), sizeof(buf)));
return ss.str();
}
}

View File

@ -20,7 +20,10 @@
#include "libhdfspp/status.h"
#include <sstream>
#include <asio/error_code.hpp>
#include <openssl/rand.h>
#include <google/protobuf/message_lite.h>
#include <google/protobuf/io/coded_stream.h>
@ -53,6 +56,10 @@ static inline void ReadDelimitedPBMessage(
std::string Base64Encode(const std::string &src);
/*
* Returns a new high-entropy client name
*/
std::string GetRandomClientName();
}
#endif

View File

@ -0,0 +1,2 @@
add_library(connection datanodeconnection.cc)
add_dependencies(connection proto)

View File

@ -0,0 +1,57 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "datanodeconnection.h"
#include "common/util.h"
namespace hdfs {
DataNodeConnection::~DataNodeConnection(){}
DataNodeConnectionImpl::~DataNodeConnectionImpl(){}
DataNodeConnectionImpl::DataNodeConnectionImpl(asio::io_service * io_service,
const ::hadoop::hdfs::DatanodeInfoProto &dn_proto,
const hadoop::common::TokenProto *token)
{
using namespace ::asio::ip;
conn_.reset(new tcp::socket(*io_service));
auto datanode_addr = dn_proto.id();
endpoints_[0] = tcp::endpoint(address::from_string(datanode_addr.ipaddr()),
datanode_addr.xferport());
uuid_ = dn_proto.id().datanodeuuid();
if (token) {
token_.reset(new hadoop::common::TokenProto());
token_->CheckTypeAndMergeFrom(*token);
}
}
void DataNodeConnectionImpl::Connect(
std::function<void(Status status, std::shared_ptr<DataNodeConnection> dn)> handler) {
// Keep the DN from being freed until we're done
auto shared_this = shared_from_this();
asio::async_connect(*conn_, endpoints_.begin(), endpoints_.end(),
[shared_this, handler](const asio::error_code &ec, std::array<asio::ip::tcp::endpoint, 1>::iterator it) {
(void)it;
handler(ToStatus(ec), shared_this); });
}
}

View File

@ -0,0 +1,66 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef LIBHDFSPP_LIB_CONNECTION_DATANODECONNECTION_H_
#define LIBHDFSPP_LIB_CONNECTION_DATANODECONNECTION_H_
#include "common/hdfs_public_api.h"
#include "common/async_stream.h"
#include "ClientNamenodeProtocol.pb.h"
#include "asio.hpp"
namespace hdfs {
class DataNodeConnection : public AsyncStream {
public:
std::string uuid_;
std::unique_ptr<hadoop::common::TokenProto> token_;
virtual ~DataNodeConnection();
virtual void Connect(std::function<void(Status status, std::shared_ptr<DataNodeConnection> dn)> handler) = 0;
};
class DataNodeConnectionImpl : public DataNodeConnection, public std::enable_shared_from_this<DataNodeConnectionImpl>{
public:
std::unique_ptr<asio::ip::tcp::socket> conn_;
std::array<asio::ip::tcp::endpoint, 1> endpoints_;
std::string uuid_;
virtual ~DataNodeConnectionImpl();
DataNodeConnectionImpl(asio::io_service * io_service, const ::hadoop::hdfs::DatanodeInfoProto &dn_proto,
const hadoop::common::TokenProto *token);
void Connect(std::function<void(Status status, std::shared_ptr<DataNodeConnection> dn)> handler) override;
void async_read_some(const MutableBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
conn_->async_read_some(buf, handler);
};
void async_write_some(const ConstBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
conn_->async_write_some(buf, handler);
}
};
}
#endif

View File

@ -1,2 +1,2 @@
add_library(fs filesystem.cc inputstream.cc bad_datanode_tracker.cc)
add_library(fs filesystem.cc filehandle.cc bad_datanode_tracker.cc)
add_dependencies(fs proto)

View File

@ -0,0 +1,240 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "filehandle.h"
#include "common/continuation/continuation.h"
#include "connection/datanodeconnection.h"
#include "reader/block_reader.h"
#include <future>
#include <tuple>
namespace hdfs {
using ::hadoop::hdfs::LocatedBlocksProto;
FileHandle::~FileHandle() {}
FileHandleImpl::FileHandleImpl(::asio::io_service *io_service, const std::string &client_name,
const std::shared_ptr<const struct FileInfo> file_info,
std::shared_ptr<BadDataNodeTracker> bad_data_nodes)
: io_service_(io_service), client_name_(client_name), file_info_(file_info),
bad_node_tracker_(bad_data_nodes), offset_(0) {
}
void FileHandleImpl::PositionRead(
void *buf, size_t nbyte, uint64_t offset,
const std::function<void(const Status &, size_t)>
&handler) {
auto callback = [this, handler](const Status &status,
const std::string &contacted_datanode,
size_t bytes_read) {
/* determine if DN gets marked bad */
if (ShouldExclude(status)) {
bad_node_tracker_->AddBadNode(contacted_datanode);
}
handler(status, bytes_read);
};
AsyncPreadSome(offset, asio::buffer(buf, nbyte), bad_node_tracker_, callback);
}
Status FileHandleImpl::PositionRead(void *buf, size_t *nbyte, off_t offset) {
auto callstate = std::make_shared<std::promise<std::tuple<Status, size_t>>>();
std::future<std::tuple<Status, size_t>> future(callstate->get_future());
/* wrap async call with promise/future to make it blocking */
auto callback = [callstate](const Status &s, size_t bytes) {
callstate->set_value(std::make_tuple(s,bytes));
};
PositionRead(buf, *nbyte, offset, callback);
/* wait for async to finish */
auto returnstate = future.get();
auto stat = std::get<0>(returnstate);
if (!stat.ok()) {
return stat;
}
*nbyte = std::get<1>(returnstate);
return stat;
}
Status FileHandleImpl::Read(void *buf, size_t *nbyte) {
Status stat = PositionRead(buf, nbyte, offset_);
if(!stat.ok()) {
return stat;
}
offset_ += *nbyte;
return Status::OK();
}
Status FileHandleImpl::Seek(off_t *offset, std::ios_base::seekdir whence) {
off_t new_offset = -1;
switch (whence) {
case std::ios_base::beg:
new_offset = *offset;
break;
case std::ios_base::cur:
new_offset = offset_ + *offset;
break;
case std::ios_base::end:
new_offset = file_info_->file_length_ + *offset;
break;
default:
/* unsupported */
return Status::InvalidArgument("Invalid Seek whence argument");
}
if(!CheckSeekBounds(new_offset)) {
return Status::InvalidArgument("Seek offset out of bounds");
}
offset_ = new_offset;
*offset = offset_;
return Status::OK();
}
/* return false if seek will be out of bounds */
bool FileHandleImpl::CheckSeekBounds(ssize_t desired_position) {
ssize_t file_length = file_info_->file_length_;
if (desired_position < 0 || desired_position >= file_length) {
return false;
}
return true;
}
/*
* Note that this method must be thread-safe w.r.t. the unsafe operations occurring
* on the FileHandle
*/
void FileHandleImpl::AsyncPreadSome(
size_t offset, const MutableBuffers &buffers,
std::shared_ptr<NodeExclusionRule> excluded_nodes,
const std::function<void(const Status &, const std::string &, size_t)> handler) {
using ::hadoop::hdfs::DatanodeInfoProto;
using ::hadoop::hdfs::LocatedBlockProto;
/**
* Note: block and chosen_dn will end up pointing to things inside
* the blocks_ vector. They shouldn't be directly deleted.
**/
auto block = std::find_if(
file_info_->blocks_.begin(), file_info_->blocks_.end(), [offset](const LocatedBlockProto &p) {
return p.offset() <= offset && offset < p.offset() + p.b().numbytes();
});
if (block == file_info_->blocks_.end()) {
handler(Status::InvalidArgument("Cannot find corresponding blocks"), "", 0);
return;
}
/**
* If user supplies a rule use it, otherwise use the tracker.
* User is responsible for making sure one of them isn't null.
**/
std::shared_ptr<NodeExclusionRule> rule =
excluded_nodes != nullptr ? excluded_nodes : bad_node_tracker_;
auto datanodes = block->locs();
auto it = std::find_if(datanodes.begin(), datanodes.end(),
[rule](const DatanodeInfoProto &dn) {
return !rule->IsBadNode(dn.id().datanodeuuid());
});
if (it == datanodes.end()) {
handler(Status::ResourceUnavailable("No datanodes available"), "", 0);
return;
}
DatanodeInfoProto &chosen_dn = *it;
uint64_t offset_within_block = offset - block->offset();
uint64_t size_within_block = std::min<uint64_t>(
block->b().numbytes() - offset_within_block, asio::buffer_size(buffers));
// This is where we will put the logic for re-using a DN connection; we can
// steal the FileHandle's dn and put it back when we're done
std::shared_ptr<DataNodeConnection> dn = CreateDataNodeConnection(io_service_, chosen_dn, nullptr /*token*/);
std::string dn_id = dn->uuid_;
std::string client_name = client_name_;
// Wrap the DN in a block reader to handle the state and logic of the
// block request protocol
std::shared_ptr<BlockReader> reader;
reader = CreateBlockReader(BlockReaderOptions(), dn);
auto read_handler = [reader, dn_id, handler](const Status & status, size_t transferred) {
handler(status, dn_id, transferred);
};
dn->Connect([handler,read_handler,block,offset_within_block,size_within_block, buffers, reader, dn_id, client_name]
(Status status, std::shared_ptr<DataNodeConnection> dn) {
(void)dn;
if (status.ok()) {
reader->AsyncReadBlock(
client_name, *block, offset_within_block,
asio::buffer(buffers, size_within_block), read_handler);
} else {
handler(status, dn_id, 0);
}
});
return;
}
std::shared_ptr<BlockReader> FileHandleImpl::CreateBlockReader(const BlockReaderOptions &options,
std::shared_ptr<DataNodeConnection> dn)
{
return std::make_shared<BlockReaderImpl>(options, dn);
}
std::shared_ptr<DataNodeConnection> FileHandleImpl::CreateDataNodeConnection(
::asio::io_service * io_service,
const ::hadoop::hdfs::DatanodeInfoProto & dn,
const hadoop::common::TokenProto * token) {
return std::make_shared<DataNodeConnectionImpl>(io_service, dn, token);
}
bool FileHandle::ShouldExclude(const Status &s) {
if (s.ok()) {
return false;
}
switch (s.code()) {
/* client side resource exhaustion */
case Status::kResourceUnavailable:
return false;
case Status::kInvalidArgument:
case Status::kUnimplemented:
case Status::kException:
default:
return true;
}
}
}

View File

@ -0,0 +1,115 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef LIBHDFSPP_LIB_FS_FILEHANDLE_H_
#define LIBHDFSPP_LIB_FS_FILEHANDLE_H_
#include "common/hdfs_public_api.h"
#include "common/async_stream.h"
#include "reader/fileinfo.h"
#include "asio.hpp"
#include "bad_datanode_tracker.h"
#include "ClientNamenodeProtocol.pb.h"
#include <mutex>
#include <iostream>
namespace hdfs {
class BlockReader;
class BlockReaderOptions;
class DataNodeConnection;
/*
* FileHandle: coordinates operations on a particular file in HDFS
*
* Threading model: not thread-safe; consumers and io_service should not call
* concurrently. PositionRead is the exceptions; they can be
* called concurrently and repeatedly.
* Lifetime: pointer returned to consumer by FileSystem::Open. Consumer is
* resonsible for freeing the object.
*/
class FileHandleImpl : public FileHandle {
public:
FileHandleImpl(::asio::io_service *io_service, const std::string &client_name,
const std::shared_ptr<const struct FileInfo> file_info,
std::shared_ptr<BadDataNodeTracker> bad_data_nodes);
/*
* [Some day reliably] Reads a particular offset into the data file.
* On error, bytes_read returns the number of bytes successfully read; on
* success, bytes_read will equal nbyte
*/
void PositionRead(
void *buf,
size_t nbyte,
uint64_t offset,
const std::function<void(const Status &status, size_t bytes_read)> &handler
) override;
/**
* Note: The nbyte argument for Read and Pread as well as the
* offset argument for Seek are in/out parameters.
*
* For Read and Pread the value referenced by nbyte should
* be set to the number of bytes to read. Before returning
* the value referenced will be set by the callee to the number
* of bytes that was successfully read.
*
* For Seek the value referenced by offset should be the number
* of bytes to shift from the specified whence position. The
* referenced value will be set to the new offset before returning.
**/
Status PositionRead(void *buf, size_t *bytes_read, off_t offset) override;
Status Read(void *buf, size_t *nbyte) override;
Status Seek(off_t *offset, std::ios_base::seekdir whence) override;
/*
* Reads some amount of data into the buffer. Will attempt to find the best
* datanode and read data from it.
*
* If an error occurs during connection or transfer, the callback will be
* called with bytes_read equal to the number of bytes successfully transferred.
* If no data nodes can be found, status will be Status::ResourceUnavailable.
*
*/
void AsyncPreadSome(size_t offset, const MutableBuffers &buffers,
std::shared_ptr<NodeExclusionRule> excluded_nodes,
const std::function<void(const Status &status,
const std::string &dn_id, size_t bytes_read)> handler);
protected:
virtual std::shared_ptr<BlockReader> CreateBlockReader(const BlockReaderOptions &options,
std::shared_ptr<DataNodeConnection> dn);
virtual std::shared_ptr<DataNodeConnection> CreateDataNodeConnection(
::asio::io_service *io_service,
const ::hadoop::hdfs::DatanodeInfoProto & dn,
const hadoop::common::TokenProto * token);
private:
::asio::io_service * const io_service_;
const std::string client_name_;
const std::shared_ptr<const struct FileInfo> file_info_;
std::shared_ptr<BadDataNodeTracker> bad_node_tracker_;
bool CheckSeekBounds(ssize_t desired_position);
off_t offset_;
};
}
#endif

View File

@ -22,7 +22,10 @@
#include <asio/ip/tcp.hpp>
#include <functional>
#include <limits>
#include <future>
#include <tuple>
namespace hdfs {
@ -32,38 +35,17 @@ static const int kNamenodeProtocolVersion = 1;
using ::asio::ip::tcp;
FileSystem::~FileSystem() {}
/*****************************************************************************
* NAMENODE OPERATIONS
****************************************************************************/
void FileSystem::New(
IoService *io_service, const Options &options, const std::string &server,
void NameNodeOperations::Connect(const std::string &server,
const std::string &service,
const std::function<void(const Status &, FileSystem *)> &handler) {
FileSystemImpl *impl = new FileSystemImpl(io_service, options);
impl->Connect(server, service, [impl, handler](const Status &stat) {
if (stat.ok()) {
handler(stat, impl);
} else {
delete impl;
handler(stat, nullptr);
}
});
}
FileSystemImpl::FileSystemImpl(IoService *io_service, const Options &options)
: io_service_(static_cast<IoServiceImpl *>(io_service)),
engine_(&io_service_->io_service(), options,
RpcEngine::GetRandomClientName(), kNamenodeProtocol,
kNamenodeProtocolVersion),
namenode_(&engine_),
bad_node_tracker_(std::make_shared<BadDataNodeTracker>()) {}
void FileSystemImpl::Connect(const std::string &server,
const std::string &service,
std::function<void(const Status &)> &&handler) {
using namespace continuation;
std::function<void(const Status &)> &handler) {
using namespace asio_continuation;
typedef std::vector<tcp::endpoint> State;
auto m = Pipeline<State>::Create();
m->Push(Resolve(&io_service_->io_service(), server, service,
m->Push(Resolve(io_service_, server, service,
std::back_inserter(m->state())))
.Push(Bind([this, m](const Continuation::Next &next) {
engine_.Connect(m->state().front(), next);
@ -76,9 +58,9 @@ void FileSystemImpl::Connect(const std::string &server,
});
}
void FileSystemImpl::Open(
const std::string &path,
const std::function<void(const Status &, InputStream *)> &handler) {
void NameNodeOperations::GetBlockLocations(const std::string & path,
std::function<void(const Status &, std::shared_ptr<const struct FileInfo>)> handler)
{
using ::hadoop::hdfs::GetBlockLocationsRequestProto;
using ::hadoop::hdfs::GetBlockLocationsResponseProto;
@ -99,10 +81,174 @@ void FileSystemImpl::Open(
[this, s](const continuation::Continuation::Next &next) {
namenode_.GetBlockLocations(&s->req, s->resp, next);
}));
m->Run([this, handler](const Status &stat, const State &s) {
handler(stat, stat.ok() ? new InputStreamImpl(this, &s.resp->locations(),
bad_node_tracker_)
if (stat.ok()) {
auto file_info = std::make_shared<struct FileInfo>();
auto locations = s.resp->locations();
file_info->file_length_ = locations.filelength();
for (const auto &block : locations.blocks()) {
file_info->blocks_.push_back(block);
}
if (locations.has_lastblock() && locations.lastblock().b().numbytes()) {
file_info->blocks_.push_back(locations.lastblock());
}
handler(stat, file_info);
} else {
handler(stat, nullptr);
}
});
}
/*****************************************************************************
* FILESYSTEM BASE CLASS
****************************************************************************/
void FileSystem::New(
IoService *io_service, const Options &options, const std::string &server,
const std::string &service,
const std::function<void(const Status &, FileSystem *)> &handler) {
FileSystemImpl *impl = new FileSystemImpl(io_service, options);
impl->Connect(server, service, [impl, handler](const Status &stat) {
if (stat.ok()) {
handler(stat, impl);
} else {
delete impl;
handler(stat, nullptr);
}
});
}
FileSystem * FileSystem::New(
IoService *io_service, const Options &options, const std::string &server,
const std::string &service) {
auto callstate = std::make_shared<std::promise<std::tuple<Status, FileSystem *>>>();
std::future<std::tuple<Status, FileSystem *>> future(callstate->get_future());
auto callback = [callstate](const Status &s, FileSystem * fs) {
callstate->set_value(std::make_tuple(s, fs));
};
New(io_service, options, server, service, callback);
/* block until promise is set */
auto returnstate = future.get();
if (std::get<0>(returnstate).ok()) {
return std::get<1>(returnstate);
} else {
return nullptr;
}
}
/*****************************************************************************
* FILESYSTEM IMPLEMENTATION
****************************************************************************/
FileSystemImpl::FileSystemImpl(IoService *&io_service, const Options &options)
: io_service_(static_cast<IoServiceImpl *>(io_service)),
nn_(&io_service_->io_service(), options,
GetRandomClientName(), kNamenodeProtocol,
kNamenodeProtocolVersion),
client_name_(GetRandomClientName())
{
// Poor man's move
io_service = nullptr;
/* spawn background threads for asio delegation */
unsigned int threads = 1 /* options.io_threads_, pending HDFS-9117 */;
for (unsigned int i = 0; i < threads; i++) {
AddWorkerThread();
}
}
FileSystemImpl::~FileSystemImpl() {
/**
* Note: IoService must be stopped before getting rid of worker threads.
* Once worker threads are joined and deleted the service can be deleted.
**/
io_service_->Stop();
worker_threads_.clear();
io_service_.reset(nullptr);
}
void FileSystemImpl::Connect(const std::string &server,
const std::string &service,
std::function<void(const Status &)> &&handler) {
/* IoService::New can return nullptr */
if (!io_service_) {
handler (Status::Error("Null IoService"));
}
nn_.Connect(server, service, handler);
}
Status FileSystemImpl::Connect(const std::string &server, const std::string &service) {
/* synchronized */
auto stat = std::make_shared<std::promise<Status>>();
std::future<Status> future = stat->get_future();
auto callback = [stat](const Status &s) {
stat->set_value(s);
};
Connect(server, service, callback);
/* block until promise is set */
auto s = future.get();
return s;
}
int FileSystemImpl::AddWorkerThread() {
auto service_task = [](IoService *service) { service->Run(); };
worker_threads_.push_back(
WorkerPtr(new std::thread(service_task, io_service_.get())));
return worker_threads_.size();
}
void FileSystemImpl::Open(
const std::string &path,
const std::function<void(const Status &, FileHandle *)> &handler) {
nn_.GetBlockLocations(path, [this, handler](const Status &stat, std::shared_ptr<const struct FileInfo> file_info) {
handler(stat, stat.ok() ? new FileHandleImpl(&io_service_->io_service(), client_name_, file_info, bad_node_tracker_)
: nullptr);
});
}
Status FileSystemImpl::Open(const std::string &path,
FileHandle **handle) {
auto callstate = std::make_shared<std::promise<std::tuple<Status, FileHandle*>>>();
std::future<std::tuple<Status, FileHandle*>> future(callstate->get_future());
/* wrap async FileSystem::Open with promise to make it a blocking call */
auto h = [callstate](const Status &s, FileHandle *is) {
callstate->set_value(std::make_tuple(s, is));
};
Open(path, h);
/* block until promise is set */
auto returnstate = future.get();
Status stat = std::get<0>(returnstate);
FileHandle *file_handle = std::get<1>(returnstate);
if (!stat.ok()) {
delete file_handle;
return stat;
}
if (!file_handle) {
return stat;
}
*handle = file_handle;
return stat;
}
}

View File

@ -18,75 +18,110 @@
#ifndef LIBHDFSPP_LIB_FS_FILESYSTEM_H_
#define LIBHDFSPP_LIB_FS_FILESYSTEM_H_
#include "filehandle.h"
#include "common/hdfs_public_api.h"
#include "common/async_stream.h"
#include "libhdfspp/hdfs.h"
#include "fs/bad_datanode_tracker.h"
#include "rpc/rpc_engine.h"
#include "reader/block_reader.h"
#include "reader/fileinfo.h"
#include "ClientNamenodeProtocol.pb.h"
#include "ClientNamenodeProtocol.hrpc.inl"
#include "asio.hpp"
#include <thread>
namespace hdfs {
class FileHandle;
class HadoopFileSystem;
/**
* NameNodeConnection: abstracts the details of communicating with a NameNode
* and the implementation of the communications protocol.
*
* Will eventually handle retry and failover.
*
* Threading model: thread-safe; all operations can be called concurrently
* Lifetime: owned by a FileSystemImpl
*/
class NameNodeOperations {
public:
NameNodeOperations(::asio::io_service *io_service, const Options &options,
const std::string &client_name, const char *protocol_name,
int protocol_version) :
io_service_(io_service),
engine_(io_service, options, client_name, protocol_name, protocol_version),
namenode_(& engine_) {}
class FileSystemImpl : public FileSystem {
public:
FileSystemImpl(IoService *io_service, const Options &options);
void Connect(const std::string &server, const std::string &service,
std::function<void(const Status &)> &&handler);
virtual void Open(const std::string &path,
const std::function<void(const Status &, InputStream *)> &
handler) override;
RpcEngine &rpc_engine() { return engine_; }
void Connect(const std::string &server,
const std::string &service,
std::function<void(const Status &)> &handler);
private:
IoServiceImpl *io_service_;
void GetBlockLocations(const std::string & path,
std::function<void(const Status &, std::shared_ptr<const struct FileInfo>)> handler);
private:
::asio::io_service * io_service_;
RpcEngine engine_;
ClientNamenodeProtocol namenode_;
std::shared_ptr<BadDataNodeTracker> bad_node_tracker_;
};
class InputStreamImpl : public InputStream {
public:
InputStreamImpl(FileSystemImpl *fs,
const ::hadoop::hdfs::LocatedBlocksProto *blocks,
std::shared_ptr<BadDataNodeTracker> tracker);
virtual void PositionRead(
void *buf, size_t nbyte, uint64_t offset,
const std::function<void(const Status &, const std::string &, size_t)> &
handler) override;
/**
* If optional_rule_override is null then use the bad_datanode_tracker. If
* non-null use the provided NodeExclusionRule to determine eligible
* datanodes.
**/
template <class MutableBufferSequence, class Handler>
void AsyncPreadSome(size_t offset, const MutableBufferSequence &buffers,
std::shared_ptr<NodeExclusionRule> excluded_nodes,
const Handler &handler);
/*
* FileSystem: The consumer's main point of interaction with the cluster as
* a whole.
*
* Initially constructed in a disconnected state; call Connect before operating
* on the FileSystem.
*
* All open files must be closed before the FileSystem is destroyed.
*
* Threading model: thread-safe for all operations
* Lifetime: pointer created for consumer who is responsible for deleting it
*/
class FileSystemImpl : public FileSystem {
public:
FileSystemImpl(IoService *&io_service, const Options &options);
~FileSystemImpl() override;
template <class BlockReaderTrait, class MutableBufferSequence, class Handler>
void AsyncReadBlock(const std::string &client_name,
const hadoop::hdfs::LocatedBlockProto &block,
const hadoop::hdfs::DatanodeInfoProto &dn, size_t offset,
const MutableBufferSequence &buffers,
const Handler &handler);
uint64_t get_file_length() const;
private:
FileSystemImpl *fs_;
unsigned long long file_length_;
std::vector<::hadoop::hdfs::LocatedBlockProto> blocks_;
template <class Reader>
struct HandshakeContinuation;
template <class Reader, class MutableBufferSequence>
struct ReadBlockContinuation;
struct RemoteBlockReaderTrait;
friend class FileHandle;
/* attempt to connect to namenode, return bad status on failure */
void Connect(const std::string &server, const std::string &service,
std::function<void(const Status &)> &&handler);
/* attempt to connect to namenode, return bad status on failure */
Status Connect(const std::string &server, const std::string &service);
virtual void Open(const std::string &path,
const std::function<void(const Status &, FileHandle *)>
&handler) override;
Status Open(const std::string &path, FileHandle **handle) override;
/* add a new thread to handle asio requests, return number of threads in pool
*/
int AddWorkerThread();
/* how many worker threads are servicing asio requests */
int WorkerThreadCount() { return worker_threads_.size(); }
private:
std::unique_ptr<IoServiceImpl> io_service_;
NameNodeOperations nn_;
const std::string client_name_;
std::shared_ptr<BadDataNodeTracker> bad_node_tracker_;
struct WorkerDeleter {
void operator()(std::thread *t) {
t->join();
delete t;
}
};
typedef std::unique_ptr<std::thread, WorkerDeleter> WorkerPtr;
std::vector<WorkerPtr> worker_threads_;
};
}
#include "inputstream_impl.h"
#endif

View File

@ -1,48 +0,0 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "filesystem.h"
namespace hdfs {
using ::hadoop::hdfs::LocatedBlocksProto;
InputStream::~InputStream() {}
InputStreamImpl::InputStreamImpl(FileSystemImpl *fs,
const LocatedBlocksProto *blocks,
std::shared_ptr<BadDataNodeTracker> tracker)
: fs_(fs), file_length_(blocks->filelength()), bad_node_tracker_(tracker) {
for (const auto &block : blocks->blocks()) {
blocks_.push_back(block);
}
if (blocks->has_lastblock() && blocks->lastblock().b().numbytes()) {
blocks_.push_back(blocks->lastblock());
}
}
void InputStreamImpl::PositionRead(
void *buf, size_t nbyte, uint64_t offset,
const std::function<void(const Status &, const std::string &, size_t)> &
handler) {
AsyncPreadSome(offset, asio::buffer(buf, nbyte), bad_node_tracker_, handler);
}
uint64_t InputStreamImpl::get_file_length() const { return file_length_; }
}

View File

@ -1,207 +0,0 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef FS_INPUTSTREAM_IMPL_H_
#define FS_INPUTSTREAM_IMPL_H_
#include "reader/block_reader.h"
#include "common/continuation/asio.h"
#include "common/continuation/protobuf.h"
#include <functional>
#include <future>
#include <type_traits>
#include <algorithm>
namespace hdfs {
struct InputStreamImpl::RemoteBlockReaderTrait {
typedef RemoteBlockReader<asio::ip::tcp::socket> Reader;
struct State {
std::unique_ptr<asio::ip::tcp::socket> conn_;
std::shared_ptr<Reader> reader_;
std::array<asio::ip::tcp::endpoint, 1> endpoints_;
size_t transferred_;
Reader *reader() { return reader_.get(); }
size_t *transferred() { return &transferred_; }
const size_t *transferred() const { return &transferred_; }
};
static continuation::Pipeline<State> *CreatePipeline(
::asio::io_service *io_service,
const ::hadoop::hdfs::DatanodeInfoProto &dn) {
using namespace ::asio::ip;
auto m = continuation::Pipeline<State>::Create();
auto &s = m->state();
s.conn_.reset(new tcp::socket(*io_service));
s.reader_ = std::make_shared<Reader>(BlockReaderOptions(), s.conn_.get());
auto datanode = dn.id();
s.endpoints_[0] = tcp::endpoint(address::from_string(datanode.ipaddr()),
datanode.xferport());
m->Push(continuation::Connect(s.conn_.get(), s.endpoints_.begin(),
s.endpoints_.end()));
return m;
}
};
template <class Reader>
struct InputStreamImpl::HandshakeContinuation : continuation::Continuation {
HandshakeContinuation(Reader *reader, const std::string &client_name,
const hadoop::common::TokenProto *token,
const hadoop::hdfs::ExtendedBlockProto *block,
uint64_t length, uint64_t offset)
: reader_(reader),
client_name_(client_name),
length_(length),
offset_(offset) {
if (token) {
token_.reset(new hadoop::common::TokenProto());
token_->CheckTypeAndMergeFrom(*token);
}
block_.CheckTypeAndMergeFrom(*block);
}
virtual void Run(const Next &next) override {
reader_->async_connect(client_name_, token_.get(), &block_, length_,
offset_, next);
}
private:
Reader *reader_;
const std::string client_name_;
std::unique_ptr<hadoop::common::TokenProto> token_;
hadoop::hdfs::ExtendedBlockProto block_;
uint64_t length_;
uint64_t offset_;
};
template <class Reader, class MutableBufferSequence>
struct InputStreamImpl::ReadBlockContinuation : continuation::Continuation {
ReadBlockContinuation(Reader *reader, MutableBufferSequence buffer,
size_t *transferred)
: reader_(reader),
buffer_(buffer),
buffer_size_(asio::buffer_size(buffer)),
transferred_(transferred) {
static_assert(!std::is_reference<MutableBufferSequence>::value,
"Buffer must not be a reference type");
}
virtual void Run(const Next &next) override {
*transferred_ = 0;
next_ = next;
OnReadData(Status::OK(), 0);
}
private:
Reader *reader_;
const MutableBufferSequence buffer_;
const size_t buffer_size_;
size_t *transferred_;
std::function<void(const Status &)> next_;
void OnReadData(const Status &status, size_t transferred) {
using std::placeholders::_1;
using std::placeholders::_2;
*transferred_ += transferred;
if (!status.ok()) {
next_(status);
} else if (*transferred_ >= buffer_size_) {
next_(status);
} else {
reader_->async_read_some(
asio::buffer(buffer_ + *transferred_, buffer_size_ - *transferred_),
std::bind(&ReadBlockContinuation::OnReadData, this, _1, _2));
}
}
};
template <class MutableBufferSequence, class Handler>
void InputStreamImpl::AsyncPreadSome(
size_t offset, const MutableBufferSequence &buffers,
std::shared_ptr<NodeExclusionRule> excluded_nodes, const Handler &handler) {
using ::hadoop::hdfs::DatanodeInfoProto;
using ::hadoop::hdfs::LocatedBlockProto;
/**
* Note: block and chosen_dn will end up pointing to things inside
* the blocks_ vector. They shouldn't be directly deleted.
**/
auto block = std::find_if(
blocks_.begin(), blocks_.end(), [offset](const LocatedBlockProto &p) {
return p.offset() <= offset && offset < p.offset() + p.b().numbytes();
});
if (block == blocks_.end()) {
handler(Status::InvalidArgument("Cannot find corresponding blocks"), "", 0);
return;
}
/**
* If user supplies a rule use it, otherwise use the tracker.
* User is responsible for making sure one of them isn't null.
**/
std::shared_ptr<NodeExclusionRule> rule =
excluded_nodes != nullptr ? excluded_nodes : bad_node_tracker_;
auto datanodes = block->locs();
auto it = std::find_if(datanodes.begin(), datanodes.end(),
[rule](const DatanodeInfoProto &dn) {
return !rule->IsBadNode(dn.id().datanodeuuid());
});
if (it == datanodes.end()) {
handler(Status::ResourceUnavailable("No datanodes available"), "", 0);
return;
}
DatanodeInfoProto *chosen_dn = &*it;
uint64_t offset_within_block = offset - block->offset();
uint64_t size_within_block = std::min<uint64_t>(
block->b().numbytes() - offset_within_block, asio::buffer_size(buffers));
AsyncReadBlock<RemoteBlockReaderTrait>(
fs_->rpc_engine().client_name(), *block, *chosen_dn, offset_within_block,
asio::buffer(buffers, size_within_block), handler);
}
template <class BlockReaderTrait, class MutableBufferSequence, class Handler>
void InputStreamImpl::AsyncReadBlock(
const std::string &client_name,
const hadoop::hdfs::LocatedBlockProto &block,
const hadoop::hdfs::DatanodeInfoProto &dn, size_t offset,
const MutableBufferSequence &buffers, const Handler &handler) {
typedef typename BlockReaderTrait::Reader Reader;
auto m =
BlockReaderTrait::CreatePipeline(&fs_->rpc_engine().io_service(), dn);
auto &s = m->state();
size_t size = asio::buffer_size(buffers);
m->Push(new HandshakeContinuation<Reader>(s.reader(), client_name, nullptr,
&block.b(), size, offset))
.Push(new ReadBlockContinuation<Reader, MutableBufferSequence>(
s.reader(), buffers, s.transferred()));
const std::string &dnid = dn.id().datanodeuuid();
m->Run([handler, dnid](const Status &status,
const typename BlockReaderTrait::State &state) {
handler(status, dnid, *state.transferred());
});
}
}
#endif

View File

@ -16,5 +16,5 @@
# limitations under the License.
#
add_library(reader remote_block_reader.cc datatransfer.cc)
add_library(reader block_reader.cc datatransfer.cc)
add_dependencies(reader proto)

View File

@ -15,18 +15,10 @@
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef LIBHDFSPP_READER_REMOTE_BLOCK_READER_IMPL_H_
#define LIBHDFSPP_READER_REMOTE_BLOCK_READER_IMPL_H_
#include "datatransfer.h"
#include "reader/block_reader.h"
#include "reader/datatransfer.h"
#include "common/continuation/continuation.h"
#include "common/continuation/asio.h"
#include "common/continuation/protobuf.h"
#include <asio/buffers_iterator.hpp>
#include <asio/streambuf.hpp>
#include <asio/write.hpp>
#include <arpa/inet.h>
#include <future>
@ -36,14 +28,31 @@ hadoop::hdfs::OpReadBlockProto
ReadBlockProto(const std::string &client_name, bool verify_checksum,
const hadoop::common::TokenProto *token,
const hadoop::hdfs::ExtendedBlockProto *block, uint64_t length,
uint64_t offset);
uint64_t offset) {
using namespace hadoop::hdfs;
using namespace hadoop::common;
BaseHeaderProto *base_h = new BaseHeaderProto();
base_h->set_allocated_block(new ExtendedBlockProto(*block));
if (token) {
base_h->set_allocated_token(new TokenProto(*token));
}
ClientOperationHeaderProto *h = new ClientOperationHeaderProto();
h->set_clientname(client_name);
h->set_allocated_baseheader(base_h);
template <class Stream>
template <class ConnectHandler>
void RemoteBlockReader<Stream>::async_connect(
const std::string &client_name, const hadoop::common::TokenProto *token,
OpReadBlockProto p;
p.set_allocated_header(h);
p.set_offset(offset);
p.set_len(length);
p.set_sendchecksums(verify_checksum);
// TODO: p.set_allocated_cachingstrategy();
return p;
}
void BlockReaderImpl::AsyncRequestBlock(
const std::string &client_name,
const hadoop::hdfs::ExtendedBlockProto *block, uint64_t length,
uint64_t offset, const ConnectHandler &handler) {
uint64_t offset, const std::function<void(Status)> &handler) {
// The total number of bytes that we need to transfer from the DN is
// the amount that the user wants (bytesToRead), plus the padding at
// the beginning in order to chunk-align. Note that the DN may elect
@ -62,18 +71,17 @@ void RemoteBlockReader<Stream>::async_connect(
s->header.insert(s->header.begin(),
{0, kDataTransferVersion, Operation::kReadBlock});
s->request = std::move(ReadBlockProto(client_name, options_.verify_checksum,
token, block, length, offset));
dn_->token_.get(), block, length, offset));
auto read_pb_message =
new continuation::ReadDelimitedPBMessageContinuation<Stream, 16384>(
stream_, &s->response);
new continuation::ReadDelimitedPBMessageContinuation<AsyncStream, 16384>(
dn_, &s->response);
m->Push(continuation::Write(stream_, asio::buffer(s->header)))
.Push(continuation::WriteDelimitedPBMessage(stream_, &s->request))
m->Push(asio_continuation::Write(dn_.get(), asio::buffer(s->header)))
.Push(asio_continuation::WriteDelimitedPBMessage(dn_, &s->request))
.Push(read_pb_message);
m->Run([this, handler, offset](const Status &status, const State &s) {
Status stat = status;
m->Run([this, handler, offset](const Status &status, const State &s) { Status stat = status;
if (stat.ok()) {
const auto &resp = s.response;
if (resp.status() == ::hadoop::hdfs::Status::SUCCESS) {
@ -90,10 +98,26 @@ void RemoteBlockReader<Stream>::async_connect(
});
}
template <class Stream>
struct RemoteBlockReader<Stream>::ReadPacketHeader
Status BlockReaderImpl::RequestBlock(
const std::string &client_name,
const hadoop::hdfs::ExtendedBlockProto *block, uint64_t length,
uint64_t offset) {
auto stat = std::make_shared<std::promise<Status>>();
std::future<Status> future(stat->get_future());
AsyncRequestBlock(client_name, block, length, offset,
[stat](const Status &status) { stat->set_value(status); });
return future.get();
}
hadoop::hdfs::OpReadBlockProto
ReadBlockProto(const std::string &client_name, bool verify_checksum,
const hadoop::common::TokenProto *token,
const hadoop::hdfs::ExtendedBlockProto *block, uint64_t length,
uint64_t offset);
struct BlockReaderImpl::ReadPacketHeader
: continuation::Continuation {
ReadPacketHeader(RemoteBlockReader<Stream> *parent) : parent_(parent) {}
ReadPacketHeader(BlockReaderImpl *parent) : parent_(parent) {}
virtual void Run(const Next &next) override {
parent_->packet_data_read_bytes_ = 0;
@ -113,7 +137,7 @@ struct RemoteBlockReader<Stream>::ReadPacketHeader
next(status);
};
asio::async_read(*parent_->stream_, asio::buffer(buf_),
asio::async_read(*parent_->dn_, asio::buffer(buf_),
std::bind(&ReadPacketHeader::CompletionHandler, this,
std::placeholders::_1, std::placeholders::_2),
handler);
@ -127,7 +151,7 @@ private:
static const size_t kHeaderLenSize = sizeof(int16_t);
static const size_t kHeaderStart = kPayloadLenSize + kHeaderLenSize;
RemoteBlockReader<Stream> *parent_;
BlockReaderImpl *parent_;
std::array<char, kMaxHeaderSize> buf_;
size_t packet_length() const {
@ -149,9 +173,8 @@ private:
}
};
template <class Stream>
struct RemoteBlockReader<Stream>::ReadChecksum : continuation::Continuation {
ReadChecksum(RemoteBlockReader<Stream> *parent) : parent_(parent) {}
struct BlockReaderImpl::ReadChecksum : continuation::Continuation {
ReadChecksum(BlockReaderImpl *parent) : parent_(parent) {}
virtual void Run(const Next &next) override {
auto parent = parent_;
@ -172,20 +195,58 @@ struct RemoteBlockReader<Stream>::ReadChecksum : continuation::Continuation {
};
parent->checksum_.resize(parent->packet_len_ - sizeof(int) -
parent->header_.datalen());
asio::async_read(*parent->stream_, asio::buffer(parent->checksum_),
asio::async_read(*parent->dn_, asio::buffer(parent->checksum_), handler);
}
private:
BlockReaderImpl *parent_;
};
struct BlockReaderImpl::ReadData : continuation::Continuation {
ReadData(BlockReaderImpl *parent,
std::shared_ptr<size_t> bytes_transferred,
const asio::mutable_buffers_1 &buf)
: parent_(parent), bytes_transferred_(bytes_transferred), buf_(buf) {
buf_.begin();
}
~ReadData() {
buf_.end();
}
virtual void Run(const Next &next) override {
auto handler =
[next, this](const asio::error_code &ec, size_t transferred) {
Status status;
if (ec) {
status = Status(ec.value(), ec.message().c_str());
}
*bytes_transferred_ += transferred;
parent_->bytes_to_read_ -= transferred;
parent_->packet_data_read_bytes_ += transferred;
if (parent_->packet_data_read_bytes_ >= parent_->header_.datalen()) {
parent_->state_ = kReadPacketHeader;
}
next(status);
};
auto data_len =
parent_->header_.datalen() - parent_->packet_data_read_bytes_;
asio::async_read(*parent_->dn_, buf_, asio::transfer_exactly(data_len),
handler);
}
private:
RemoteBlockReader<Stream> *parent_;
BlockReaderImpl *parent_;
std::shared_ptr<size_t> bytes_transferred_;
const asio::mutable_buffers_1 buf_;
};
template <class Stream>
struct RemoteBlockReader<Stream>::ReadPadding : continuation::Continuation {
ReadPadding(RemoteBlockReader<Stream> *parent)
struct BlockReaderImpl::ReadPadding : continuation::Continuation {
ReadPadding(BlockReaderImpl *parent)
: parent_(parent), padding_(parent->chunk_padding_bytes_),
bytes_transferred_(std::make_shared<size_t>(0)),
read_data_(new ReadData<asio::mutable_buffers_1>(
read_data_(new ReadData(
parent, bytes_transferred_, asio::buffer(padding_))) {}
virtual void Run(const Next &next) override {
@ -207,7 +268,7 @@ struct RemoteBlockReader<Stream>::ReadPadding : continuation::Continuation {
}
private:
RemoteBlockReader<Stream> *parent_;
BlockReaderImpl *parent_;
std::vector<char> padding_;
std::shared_ptr<size_t> bytes_transferred_;
std::shared_ptr<continuation::Continuation> read_data_;
@ -215,45 +276,9 @@ private:
ReadPadding &operator=(const ReadPadding &) = delete;
};
template <class Stream>
template <class MutableBufferSequence>
struct RemoteBlockReader<Stream>::ReadData : continuation::Continuation {
ReadData(RemoteBlockReader<Stream> *parent,
std::shared_ptr<size_t> bytes_transferred,
const MutableBufferSequence &buf)
: parent_(parent), bytes_transferred_(bytes_transferred), buf_(buf) {}
virtual void Run(const Next &next) override {
auto handler =
[next, this](const asio::error_code &ec, size_t transferred) {
Status status;
if (ec) {
status = Status(ec.value(), ec.message().c_str());
}
*bytes_transferred_ += transferred;
parent_->bytes_to_read_ -= transferred;
parent_->packet_data_read_bytes_ += transferred;
if (parent_->packet_data_read_bytes_ >= parent_->header_.datalen()) {
parent_->state_ = kReadPacketHeader;
}
next(status);
};
auto data_len =
parent_->header_.datalen() - parent_->packet_data_read_bytes_;
async_read(*parent_->stream_, buf_, asio::transfer_exactly(data_len),
handler);
}
private:
RemoteBlockReader<Stream> *parent_;
std::shared_ptr<size_t> bytes_transferred_;
MutableBufferSequence buf_;
};
template <class Stream>
struct RemoteBlockReader<Stream>::AckRead : continuation::Continuation {
AckRead(RemoteBlockReader<Stream> *parent) : parent_(parent) {}
struct BlockReaderImpl::AckRead : continuation::Continuation {
AckRead(BlockReaderImpl *parent) : parent_(parent) {}
virtual void Run(const Next &next) override {
if (parent_->bytes_to_read_ > 0) {
@ -268,25 +293,24 @@ struct RemoteBlockReader<Stream>::AckRead : continuation::Continuation {
: hadoop::hdfs::Status::SUCCESS);
m->Push(
continuation::WriteDelimitedPBMessage(parent_->stream_, &m->state()));
continuation::WriteDelimitedPBMessage(parent_->dn_, &m->state()));
m->Run([this, next](const Status &status,
const hadoop::hdfs::ClientReadStatusProto &) {
if (status.ok()) {
parent_->state_ = RemoteBlockReader<Stream>::kFinished;
parent_->state_ = BlockReaderImpl::kFinished;
}
next(status);
});
}
private:
RemoteBlockReader<Stream> *parent_;
BlockReaderImpl *parent_;
};
template <class Stream>
template <class MutableBufferSequence, class ReadHandler>
void RemoteBlockReader<Stream>::async_read_some(
const MutableBufferSequence &buffers, const ReadHandler &handler) {
void BlockReaderImpl::AsyncReadPacket(
const MutableBuffers &buffers,
const std::function<void(const Status &, size_t bytes_transferred)> &handler) {
assert(state_ != kOpen && "Not connected");
struct State {
@ -298,7 +322,7 @@ void RemoteBlockReader<Stream>::async_read_some(
m->Push(new ReadPacketHeader(this))
.Push(new ReadChecksum(this))
.Push(new ReadPadding(this))
.Push(new ReadData<MutableBufferSequence>(
.Push(new ReadData(
this, m->state().bytes_transferred, buffers))
.Push(new AckRead(this));
@ -308,15 +332,14 @@ void RemoteBlockReader<Stream>::async_read_some(
});
}
template <class Stream>
template <class MutableBufferSequence>
size_t
RemoteBlockReader<Stream>::read_some(const MutableBufferSequence &buffers,
BlockReaderImpl::ReadPacket(const MutableBuffers &buffers,
Status *status) {
size_t transferred = 0;
auto done = std::make_shared<std::promise<void>>();
auto future = done->get_future();
async_read_some(buffers,
AsyncReadPacket(buffers,
[status, &transferred, done](const Status &stat, size_t t) {
*status = stat;
transferred = t;
@ -326,17 +349,85 @@ RemoteBlockReader<Stream>::read_some(const MutableBufferSequence &buffers,
return transferred;
}
template <class Stream>
Status RemoteBlockReader<Stream>::connect(
const std::string &client_name, const hadoop::common::TokenProto *token,
const hadoop::hdfs::ExtendedBlockProto *block, uint64_t length,
uint64_t offset) {
auto stat = std::make_shared<std::promise<Status>>();
std::future<Status> future(stat->get_future());
async_connect(client_name, token, block, length, offset,
[stat](const Status &status) { stat->set_value(status); });
return future.get();
}
struct BlockReaderImpl::RequestBlockContinuation : continuation::Continuation {
RequestBlockContinuation(BlockReader *reader, const std::string &client_name,
const hadoop::hdfs::ExtendedBlockProto *block,
uint64_t length, uint64_t offset)
: reader_(reader), client_name_(client_name), length_(length),
offset_(offset) {
block_.CheckTypeAndMergeFrom(*block);
}
virtual void Run(const Next &next) override {
reader_->AsyncRequestBlock(client_name_, &block_, length_,
offset_, next);
}
private:
BlockReader *reader_;
const std::string client_name_;
hadoop::hdfs::ExtendedBlockProto block_;
uint64_t length_;
uint64_t offset_;
};
struct BlockReaderImpl::ReadBlockContinuation : continuation::Continuation {
ReadBlockContinuation(BlockReader *reader, MutableBuffers buffer,
size_t *transferred)
: reader_(reader), buffer_(buffer),
buffer_size_(asio::buffer_size(buffer)), transferred_(transferred) {
}
virtual void Run(const Next &next) override {
*transferred_ = 0;
next_ = next;
OnReadData(Status::OK(), 0);
}
private:
BlockReader *reader_;
const MutableBuffers buffer_;
const size_t buffer_size_;
size_t *transferred_;
std::function<void(const Status &)> next_;
void OnReadData(const Status &status, size_t transferred) {
using std::placeholders::_1;
using std::placeholders::_2;
*transferred_ += transferred;
if (!status.ok()) {
next_(status);
} else if (*transferred_ >= buffer_size_) {
next_(status);
} else {
reader_->AsyncReadPacket(
asio::buffer(buffer_ + *transferred_, buffer_size_ - *transferred_),
std::bind(&ReadBlockContinuation::OnReadData, this, _1, _2));
}
}
};
void BlockReaderImpl::AsyncReadBlock(
const std::string & client_name,
const hadoop::hdfs::LocatedBlockProto &block,
size_t offset,
const MutableBuffers &buffers,
const std::function<void(const Status &, size_t)> handler) {
auto m = continuation::Pipeline<size_t>::Create();
size_t * bytesTransferred = &m->state();
size_t size = asio::buffer_size(buffers);
m->Push(new RequestBlockContinuation(this, client_name,
&block.b(), size, offset))
.Push(new ReadBlockContinuation(this, buffers, bytesTransferred));
m->Run([handler] (const Status &status,
const size_t totalBytesTransferred) {
handler(status, totalBytesTransferred);
});
}
#endif
}

View File

@ -19,7 +19,9 @@
#define BLOCK_READER_H_
#include "libhdfspp/status.h"
#include "common/async_stream.h"
#include "datatransfer.pb.h"
#include "connection/datanodeconnection.h"
#include <memory>
@ -55,38 +57,73 @@ struct BlockReaderOptions {
: verify_checksum(true), encryption_scheme(EncryptionScheme::kNone) {}
};
template <class Stream>
class RemoteBlockReader
: public std::enable_shared_from_this<RemoteBlockReader<Stream>> {
/**
* Handles the operational state of request and reading a block (or portion of
* a block) from a DataNode.
*
* Threading model: not thread-safe.
* Lifecycle: should be created, used for a single read, then freed.
*/
class BlockReader {
public:
explicit RemoteBlockReader(const BlockReaderOptions &options, Stream *stream)
: stream_(stream), state_(kOpen), options_(options),
virtual void AsyncReadBlock(
const std::string & client_name,
const hadoop::hdfs::LocatedBlockProto &block, size_t offset,
const MutableBuffers &buffers,
const std::function<void(const Status &, size_t)> handler) = 0;
virtual void AsyncReadPacket(
const MutableBuffers &buffers,
const std::function<void(const Status &, size_t bytes_transferred)> &handler) = 0;
virtual void AsyncRequestBlock(
const std::string &client_name,
const hadoop::hdfs::ExtendedBlockProto *block,
uint64_t length,
uint64_t offset,
const std::function<void(Status)> &handler) = 0;
};
class BlockReaderImpl
: public BlockReader, public std::enable_shared_from_this<BlockReaderImpl> {
public:
explicit BlockReaderImpl(const BlockReaderOptions &options, std::shared_ptr<DataNodeConnection> dn)
: dn_(dn), state_(kOpen), options_(options),
chunk_padding_bytes_(0) {}
template <class MutableBufferSequence, class ReadHandler>
void async_read_some(const MutableBufferSequence &buffers,
const ReadHandler &handler);
virtual void AsyncReadPacket(
const MutableBuffers &buffers,
const std::function<void(const Status &, size_t bytes_transferred)> &handler) override;
template <class MutableBufferSequence>
size_t read_some(const MutableBufferSequence &buffers, Status *status);
virtual void AsyncRequestBlock(
const std::string &client_name,
const hadoop::hdfs::ExtendedBlockProto *block,
uint64_t length,
uint64_t offset,
const std::function<void(Status)> &handler) override;
Status connect(const std::string &client_name,
const hadoop::common::TokenProto *token,
const hadoop::hdfs::ExtendedBlockProto *block, uint64_t length,
virtual void AsyncReadBlock(
const std::string & client_name,
const hadoop::hdfs::LocatedBlockProto &block, size_t offset,
const MutableBuffers &buffers,
const std::function<void(const Status &, size_t)> handler) override;
size_t ReadPacket(const MutableBuffers &buffers, Status *status);
Status RequestBlock(
const std::string &client_name,
const hadoop::hdfs::ExtendedBlockProto *block,
uint64_t length,
uint64_t offset);
template <class ConnectHandler>
void async_connect(const std::string &client_name,
const hadoop::common::TokenProto *token,
const hadoop::hdfs::ExtendedBlockProto *block,
uint64_t length, uint64_t offset,
const ConnectHandler &handler);
private:
struct RequestBlockContinuation;
struct ReadBlockContinuation;
struct ReadPacketHeader;
struct ReadChecksum;
struct ReadPadding;
template <class MutableBufferSequence> struct ReadData;
struct ReadData;
struct AckRead;
enum State {
kOpen,
@ -97,7 +134,7 @@ private:
kFinished,
};
Stream *stream_;
std::shared_ptr<DataNodeConnection> dn_;
hadoop::hdfs::PacketHeaderProto header_;
State state_;
BlockReaderOptions options_;
@ -109,6 +146,4 @@ private:
};
}
#include "remote_block_reader_impl.h"
#endif

View File

@ -19,6 +19,10 @@
#define LIB_READER_DATA_TRANSFER_H_
#include "common/sasl_authenticator.h"
#include "common/async_stream.h"
#include "connection/datanodeconnection.h"
#include <memory>
namespace hdfs {
@ -32,26 +36,32 @@ enum Operation {
kReadBlock = 81,
};
template <class Stream> class DataTransferSaslStream {
template <class Stream> class DataTransferSaslStream : public DataNodeConnection {
public:
DataTransferSaslStream(Stream *stream, const std::string &username,
DataTransferSaslStream(std::shared_ptr<Stream> stream, const std::string &username,
const std::string &password)
: stream_(stream), authenticator_(username, password) {}
template <class Handler> void Handshake(const Handler &next);
template <class MutableBufferSequence, class ReadHandler>
void async_read_some(const MutableBufferSequence &buffers,
ReadHandler &&handler);
void async_read_some(const MutableBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
stream_->async_read_some(buf, handler);
}
template <class ConstBufferSequence, class WriteHandler>
void async_write_some(const ConstBufferSequence &buffers,
WriteHandler &&handler);
void async_write_some(const ConstBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
stream_->async_write_some(buf, handler);
}
void Connect(std::function<void(Status status, std::shared_ptr<DataNodeConnection> dn)> handler) override
{(void)handler; /*TODO: Handshaking goes here*/};
private:
DataTransferSaslStream(const DataTransferSaslStream &) = delete;
DataTransferSaslStream &operator=(const DataTransferSaslStream &) = delete;
Stream *stream_;
std::shared_ptr<Stream> stream_;
DigestMD5Authenticator authenticator_;
struct ReadSaslMessage;
struct Authenticator;

View File

@ -70,7 +70,7 @@ private:
template <class Stream>
struct DataTransferSaslStream<Stream>::ReadSaslMessage
: continuation::Continuation {
ReadSaslMessage(Stream *stream, std::string *data)
ReadSaslMessage(std::shared_ptr<Stream> stream, std::string *data)
: stream_(stream), data_(data), read_pb_(stream, &resp_) {}
virtual void Run(const Next &next) override {
@ -87,7 +87,7 @@ struct DataTransferSaslStream<Stream>::ReadSaslMessage
}
private:
Stream *stream_;
std::shared_ptr<Stream> stream_;
std::string *data_;
hadoop::hdfs::DataTransferEncryptorMessageProto resp_;
continuation::ReadDelimitedPBMessageContinuation<Stream, 1024> read_pb_;
@ -97,7 +97,7 @@ template <class Stream>
template <class Handler>
void DataTransferSaslStream<Stream>::Handshake(const Handler &next) {
using ::hadoop::hdfs::DataTransferEncryptorMessageProto;
using ::hdfs::continuation::Write;
using ::hdfs::asio_continuation::Write;
using ::hdfs::continuation::WriteDelimitedPBMessage;
static const int kMagicNumber = htonl(kDataTransferSasl);
@ -109,7 +109,7 @@ void DataTransferSaslStream<Stream>::Handshake(const Handler &next) {
std::string resp0;
DataTransferEncryptorMessageProto req1;
std::string resp1;
Stream *stream;
std::shared_ptr<Stream> stream;
};
auto m = continuation::Pipeline<State>::Create();
State *s = &m->state();
@ -117,7 +117,7 @@ void DataTransferSaslStream<Stream>::Handshake(const Handler &next) {
DataTransferSaslStreamUtil::PrepareInitialHandshake(&s->req0);
m->Push(Write(stream_, kMagicNumberBuffer))
m->Push(Write(stream_.get(), kMagicNumberBuffer))
.Push(WriteDelimitedPBMessage(stream_, &s->req0))
.Push(new ReadSaslMessage(stream_, &s->resp0))
.Push(new Authenticator(&authenticator_, &s->resp0, &s->req1))
@ -126,19 +126,6 @@ void DataTransferSaslStream<Stream>::Handshake(const Handler &next) {
m->Run([next](const Status &status, const State &) { next(status); });
}
template <class Stream>
template <class MutableBufferSequence, class ReadHandler>
void DataTransferSaslStream<Stream>::async_read_some(
const MutableBufferSequence &buffers, ReadHandler &&handler) {
stream_->async_read_some(buffers, handler);
}
template <class Stream>
template <typename ConstBufferSequence, typename WriteHandler>
void DataTransferSaslStream<Stream>::async_write_some(
const ConstBufferSequence &buffers, WriteHandler &&handler) {
stream_->async_write_some(buffers, handler);
}
}
#endif

View File

@ -0,0 +1,36 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef LIB_READER_FILEINFO_H_
#define LIB_READER_FILEINFO_H_
#include "ClientNamenodeProtocol.pb.h"
namespace hdfs {
/**
* Information that is assumed to be unchanging about a file for the duration of
* the operations.
*/
struct FileInfo {
unsigned long long file_length_;
std::vector<::hadoop::hdfs::LocatedBlockProto> blocks_;
};
}
#endif

View File

@ -1,46 +0,0 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "block_reader.h"
namespace hdfs {
hadoop::hdfs::OpReadBlockProto
ReadBlockProto(const std::string &client_name, bool verify_checksum,
const hadoop::common::TokenProto *token,
const hadoop::hdfs::ExtendedBlockProto *block, uint64_t length,
uint64_t offset) {
using namespace hadoop::hdfs;
using namespace hadoop::common;
BaseHeaderProto *base_h = new BaseHeaderProto();
base_h->set_allocated_block(new ExtendedBlockProto(*block));
if (token) {
base_h->set_allocated_token(new TokenProto(*token));
}
ClientOperationHeaderProto *h = new ClientOperationHeaderProto();
h->set_clientname(client_name);
h->set_allocated_baseheader(base_h);
OpReadBlockProto p;
p.set_allocated_header(h);
p.set_offset(offset);
p.set_len(length);
p.set_sendchecksums(verify_checksum);
// TODO: p.set_allocated_cachingstrategy();
return p;
}
}

View File

@ -19,9 +19,6 @@
#include "rpc_connection.h"
#include "common/util.h"
#include <openssl/rand.h>
#include <sstream>
#include <future>
namespace hdfs {
@ -83,15 +80,4 @@ Status RpcEngine::RawRpc(const std::string &method_name, const std::string &req,
return future.get();
}
std::string RpcEngine::GetRandomClientName() {
unsigned char buf[6] = {
0,
};
RAND_pseudo_bytes(buf, sizeof(buf));
std::stringstream ss;
ss << "libhdfs++_"
<< Base64Encode(std::string(reinterpret_cast<char *>(buf), sizeof(buf)));
return ss.str();
}
}

View File

@ -32,7 +32,7 @@ include_directories(
${LIBHDFS_SRC_DIR}
${OS_DIR}
)
add_library(hdfspp_test_shim_static STATIC hdfs_shim.c libhdfs_wrapper.c libhdfspp_wrapper.cc ${LIBHDFSPP_BINDING_C}/hdfs_cpp.cc)
add_library(hdfspp_test_shim_static STATIC hdfs_shim.c libhdfs_wrapper.c libhdfspp_wrapper.cc ${LIBHDFSPP_BINDING_C}/hdfs.cc)
add_library(test_common OBJECT mock_connection.cc)
@ -44,24 +44,20 @@ protobuf_generate_cpp(PROTO_TEST_SRCS PROTO_TEST_HDRS
)
add_executable(remote_block_reader_test remote_block_reader_test.cc $<TARGET_OBJECTS:test_common>)
target_link_libraries(remote_block_reader_test reader proto common ${PROTOBUF_LIBRARIES} ${OPENSSL_LIBRARIES} gmock_main ${CMAKE_THREAD_LIBS_INIT})
target_link_libraries(remote_block_reader_test reader proto common connection ${PROTOBUF_LIBRARIES} ${OPENSSL_LIBRARIES} gmock_main ${CMAKE_THREAD_LIBS_INIT})
add_test(remote_block_reader remote_block_reader_test)
add_executable(sasl_digest_md5_test sasl_digest_md5_test.cc)
target_link_libraries(sasl_digest_md5_test common ${OPENSSL_LIBRARIES} gmock_main ${CMAKE_THREAD_LIBS_INIT})
add_test(sasl_digest_md5 sasl_digest_md5_test)
add_executable(inputstream_test inputstream_test.cc)
target_link_libraries(inputstream_test fs rpc reader proto common ${PROTOBUF_LIBRARIES} ${OPENSSL_LIBRARIES} gmock_main ${CMAKE_THREAD_LIBS_INIT})
add_test(inputstream inputstream_test)
include_directories(${CMAKE_CURRENT_BINARY_DIR})
add_executable(rpc_engine_test rpc_engine_test.cc ${PROTO_TEST_SRCS} ${PROTO_TEST_HDRS} $<TARGET_OBJECTS:test_common>)
target_link_libraries(rpc_engine_test rpc proto common ${PROTOBUF_LIBRARIES} ${OPENSSL_LIBRARIES} gmock_main ${CMAKE_THREAD_LIBS_INIT})
add_test(rpc_engine rpc_engine_test)
add_executable(bad_datanode_test bad_datanode_test.cc)
target_link_libraries(bad_datanode_test rpc reader proto fs bindings_c rpc proto common reader ${PROTOBUF_LIBRARIES} ${OPENSSL_LIBRARIES} gmock_main ${CMAKE_THREAD_LIBS_INIT})
target_link_libraries(bad_datanode_test rpc reader proto fs bindings_c rpc proto common reader connection ${PROTOBUF_LIBRARIES} ${OPENSSL_LIBRARIES} gmock_main ${CMAKE_THREAD_LIBS_INIT})
add_test(bad_datanode bad_datanode_test)
add_executable(node_exclusion_test node_exclusion_test.cc)
@ -73,5 +69,5 @@ target_link_libraries(configuration_test common gmock_main ${CMAKE_THREAD_LIBS_I
add_test(configuration configuration_test)
build_libhdfs_test(libhdfs_threaded hdfspp_test_shim_static expect.c test_libhdfs_threaded.c ${OS_DIR}/thread.c)
link_libhdfs_test(libhdfs_threaded hdfspp_test_shim_static fs reader rpc proto common ${PROTOBUF_LIBRARIES} ${OPENSSL_LIBRARIES} native_mini_dfs ${JAVA_JVM_LIBRARY})
link_libhdfs_test(libhdfs_threaded hdfspp_test_shim_static fs reader rpc proto common connection ${PROTOBUF_LIBRARIES} ${OPENSSL_LIBRARIES} native_mini_dfs ${JAVA_JVM_LIBRARY})
add_libhdfs_test(libhdfs_threaded hdfspp_test_shim_static)

View File

@ -19,6 +19,8 @@
#include "fs/filesystem.h"
#include "fs/bad_datanode_tracker.h"
#include "common/util.h"
#include <gmock/gmock.h>
using hadoop::common::TokenProto;
@ -34,70 +36,140 @@ using ::testing::Return;
using namespace hdfs;
class MockReader {
public:
virtual ~MockReader() {}
class MockReader : public BlockReader {
public:
MOCK_METHOD2(
async_read_some,
AsyncReadPacket,
void(const asio::mutable_buffers_1 &,
const std::function<void(const Status &, size_t transferred)> &));
MOCK_METHOD6(async_connect,
void(const std::string &, TokenProto *, ExtendedBlockProto *,
uint64_t, uint64_t,
const std::function<void(const Status &)> &));
MOCK_METHOD5(AsyncRequestBlock,
void(const std::string &client_name,
const hadoop::hdfs::ExtendedBlockProto *block,
uint64_t length, uint64_t offset,
const std::function<void(Status)> &handler));
MOCK_METHOD5(AsyncReadBlock, void(
const std::string & client_name,
const hadoop::hdfs::LocatedBlockProto &block,
size_t offset,
const MutableBuffers &buffers,
const std::function<void(const Status &, size_t)> handler));
};
template <class Trait>
struct MockBlockReaderTrait {
typedef MockReader Reader;
struct State {
MockReader reader_;
size_t transferred_;
Reader *reader() { return &reader_; }
size_t *transferred() { return &transferred_; }
const size_t *transferred() const { return &transferred_; }
};
class MockDNConnection : public DataNodeConnection, public std::enable_shared_from_this<MockDNConnection> {
void Connect(std::function<void(Status status, std::shared_ptr<DataNodeConnection> dn)> handler) override {
handler(Status::OK(), shared_from_this());
}
static continuation::Pipeline<State> *CreatePipeline(
::asio::io_service *, const DatanodeInfoProto &) {
auto m = continuation::Pipeline<State>::Create();
*m->state().transferred() = 0;
Trait::InitializeMockReader(m->state().reader());
return m;
void async_read_some(const MutableBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
(void)buf;
handler(asio::error::fault, 0);
}
void async_write_some(const ConstBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
(void)buf;
handler(asio::error::fault, 0);
}
};
TEST(BadDataNodeTest, RecoverableError) {
LocatedBlocksProto blocks;
LocatedBlockProto block;
DatanodeInfoProto dn;
class PartialMockFileHandle : public FileHandleImpl {
using FileHandleImpl::FileHandleImpl;
public:
std::shared_ptr<MockReader> mock_reader_ = std::make_shared<MockReader>();
protected:
std::shared_ptr<BlockReader> CreateBlockReader(const BlockReaderOptions &options,
std::shared_ptr<DataNodeConnection> dn) override
{
(void) options; (void) dn;
assert(mock_reader_);
return mock_reader_;
}
std::shared_ptr<DataNodeConnection> CreateDataNodeConnection(
::asio::io_service *io_service,
const ::hadoop::hdfs::DatanodeInfoProto & dn,
const hadoop::common::TokenProto * token) override {
(void) io_service; (void) dn; (void) token;
return std::make_shared<MockDNConnection>();
}
};
TEST(BadDataNodeTest, TestNoNodes) {
auto file_info = std::make_shared<struct FileInfo>();
file_info->blocks_.push_back(LocatedBlockProto());
LocatedBlockProto & block = file_info->blocks_[0];
ExtendedBlockProto *b = block.mutable_b();
b->set_poolid("");
b->set_blockid(1);
b->set_generationstamp(1);
b->set_numbytes(4096);
// Set up the one block to have one datanode holding it
DatanodeInfoProto *di = block.add_locs();
DatanodeIDProto *dnid = di->mutable_id();
dnid->set_datanodeuuid("foo");
char buf[4096] = {
0,
};
IoServiceImpl io_service;
Options default_options;
FileSystemImpl fs(&io_service, default_options);
auto tracker = std::make_shared<BadDataNodeTracker>();
InputStreamImpl is(&fs, &blocks, tracker);
auto bad_node_tracker = std::make_shared<BadDataNodeTracker>();
bad_node_tracker->AddBadNode("foo");
PartialMockFileHandle is(&io_service.io_service(), GetRandomClientName(), file_info, bad_node_tracker);
Status stat;
size_t read = 0;
struct Trait {
static void InitializeMockReader(MockReader *reader) {
EXPECT_CALL(*reader, async_connect(_, _, _, _, _, _))
.WillOnce(InvokeArgument<5>(Status::OK()));
EXPECT_CALL(*reader, async_read_some(_, _))
// resource unavailable error
.WillOnce(InvokeArgument<1>(
Status::ResourceUnavailable(
"Unable to get some resource, try again later"),
sizeof(buf)));
}
// Exclude the one datanode with the data
is.AsyncPreadSome(0, asio::buffer(buf, sizeof(buf)), nullptr,
[&stat, &read](const Status &status, const std::string &, size_t transferred) {
stat = status;
read = transferred;
});
// Should fail with no resource available
ASSERT_EQ(static_cast<int>(std::errc::resource_unavailable_try_again), stat.code());
ASSERT_EQ(0UL, read);
}
TEST(BadDataNodeTest, RecoverableError) {
auto file_info = std::make_shared<struct FileInfo>();
file_info->blocks_.push_back(LocatedBlockProto());
LocatedBlockProto & block = file_info->blocks_[0];
ExtendedBlockProto *b = block.mutable_b();
b->set_poolid("");
b->set_blockid(1);
b->set_generationstamp(1);
b->set_numbytes(4096);
// Set up the one block to have one datanode holding it
DatanodeInfoProto *di = block.add_locs();
DatanodeIDProto *dnid = di->mutable_id();
dnid->set_datanodeuuid("foo");
char buf[4096] = {
0,
};
IoServiceImpl io_service;
auto tracker = std::make_shared<BadDataNodeTracker>();
PartialMockFileHandle is(&io_service.io_service(), GetRandomClientName(), file_info, tracker);
Status stat;
size_t read = 0;
EXPECT_CALL(*is.mock_reader_, AsyncReadBlock(_,_,_,_,_))
// resource unavailable error
.WillOnce(InvokeArgument<4>(
Status::ResourceUnavailable("Unable to get some resource, try again later"), 0));
is.AsyncReadBlock<MockBlockReaderTrait<Trait>>(
"client", block, dn, 0, asio::buffer(buf, sizeof(buf)),
is.AsyncPreadSome(
0, asio::buffer(buf, sizeof(buf)), nullptr,
[&stat, &read](const Status &status, const std::string &,
size_t transferred) {
stat = status;
@ -108,7 +180,7 @@ TEST(BadDataNodeTest, RecoverableError) {
std::string failing_dn = "id_of_bad_datanode";
if (!stat.ok()) {
if (InputStream::ShouldExclude(stat)) {
if (FileHandle::ShouldExclude(stat)) {
tracker->AddBadNode(failing_dn);
}
}
@ -117,35 +189,37 @@ TEST(BadDataNodeTest, RecoverableError) {
}
TEST(BadDataNodeTest, InternalError) {
LocatedBlocksProto blocks;
LocatedBlockProto block;
DatanodeInfoProto dn;
auto file_info = std::make_shared<struct FileInfo>();
file_info->blocks_.push_back(LocatedBlockProto());
LocatedBlockProto & block = file_info->blocks_[0];
ExtendedBlockProto *b = block.mutable_b();
b->set_poolid("");
b->set_blockid(1);
b->set_generationstamp(1);
b->set_numbytes(4096);
// Set up the one block to have one datanode holding it
DatanodeInfoProto *di = block.add_locs();
DatanodeIDProto *dnid = di->mutable_id();
dnid->set_datanodeuuid("foo");
char buf[4096] = {
0,
};
IoServiceImpl io_service;
Options default_options;
auto tracker = std::make_shared<BadDataNodeTracker>();
FileSystemImpl fs(&io_service, default_options);
InputStreamImpl is(&fs, &blocks, tracker);
PartialMockFileHandle is(&io_service.io_service(), GetRandomClientName(), file_info, tracker);
Status stat;
size_t read = 0;
struct Trait {
static void InitializeMockReader(MockReader *reader) {
EXPECT_CALL(*reader, async_connect(_, _, _, _, _, _))
.WillOnce(InvokeArgument<5>(Status::OK()));
EXPECT_CALL(*reader, async_read_some(_, _))
// something bad happened on the DN, calling again isn't going to help
.WillOnce(
InvokeArgument<1>(Status::Exception("server_explosion_exception",
EXPECT_CALL(*is.mock_reader_, AsyncReadBlock(_,_,_,_,_))
// resource unavailable error
.WillOnce(InvokeArgument<4>(
Status::Exception("server_explosion_exception",
"the server exploded"),
sizeof(buf)));
}
};
is.AsyncReadBlock<MockBlockReaderTrait<Trait>>(
"client", block, dn, 0, asio::buffer(buf, sizeof(buf)),
is.AsyncPreadSome(
0, asio::buffer(buf, sizeof(buf)), nullptr,
[&stat, &read](const Status &status, const std::string &,
size_t transferred) {
stat = status;
@ -156,7 +230,7 @@ TEST(BadDataNodeTest, InternalError) {
std::string failing_dn = "id_of_bad_datanode";
if (!stat.ok()) {
if (InputStream::ShouldExclude(stat)) {
if (FileHandle::ShouldExclude(stat)) {
tracker->AddBadNode(failing_dn);
}
}

View File

@ -1,232 +0,0 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "fs/filesystem.h"
#include "fs/bad_datanode_tracker.h"
#include <gmock/gmock.h>
using hadoop::common::TokenProto;
using hadoop::hdfs::DatanodeInfoProto;
using hadoop::hdfs::DatanodeIDProto;
using hadoop::hdfs::ExtendedBlockProto;
using hadoop::hdfs::LocatedBlockProto;
using hadoop::hdfs::LocatedBlocksProto;
using ::testing::_;
using ::testing::InvokeArgument;
using ::testing::Return;
using namespace hdfs;
namespace hdfs {
class MockReader {
public:
virtual ~MockReader() {}
MOCK_METHOD2(
async_read_some,
void(const asio::mutable_buffers_1 &,
const std::function<void(const Status &, size_t transferred)> &));
MOCK_METHOD6(async_connect,
void(const std::string &, TokenProto *, ExtendedBlockProto *,
uint64_t, uint64_t,
const std::function<void(const Status &)> &));
};
template <class Trait>
struct MockBlockReaderTrait {
typedef MockReader Reader;
struct State {
MockReader reader_;
size_t transferred_;
Reader *reader() { return &reader_; }
size_t *transferred() { return &transferred_; }
const size_t *transferred() const { return &transferred_; }
};
static continuation::Pipeline<State> *CreatePipeline(
::asio::io_service *, const DatanodeInfoProto &) {
auto m = continuation::Pipeline<State>::Create();
*m->state().transferred() = 0;
Trait::InitializeMockReader(m->state().reader());
return m;
}
};
}
TEST(InputStreamTest, TestReadSingleTrunk) {
LocatedBlocksProto blocks;
LocatedBlockProto block;
DatanodeInfoProto dn;
char buf[4096] = {
0,
};
IoServiceImpl io_service;
Options options;
FileSystemImpl fs(&io_service, options);
InputStreamImpl is(&fs, &blocks, std::make_shared<BadDataNodeTracker>());
Status stat;
size_t read = 0;
struct Trait {
static void InitializeMockReader(MockReader *reader) {
EXPECT_CALL(*reader, async_connect(_, _, _, _, _, _))
.WillOnce(InvokeArgument<5>(Status::OK()));
EXPECT_CALL(*reader, async_read_some(_, _))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf)));
}
};
is.AsyncReadBlock<MockBlockReaderTrait<Trait>>(
"client", block, dn, 0, asio::buffer(buf, sizeof(buf)),
[&stat, &read](const Status &status, const std::string &,
size_t transferred) {
stat = status;
read = transferred;
});
ASSERT_TRUE(stat.ok());
ASSERT_EQ(sizeof(buf), read);
read = 0;
}
TEST(InputStreamTest, TestReadMultipleTrunk) {
LocatedBlocksProto blocks;
LocatedBlockProto block;
DatanodeInfoProto dn;
char buf[4096] = {
0,
};
IoServiceImpl io_service;
Options options;
FileSystemImpl fs(&io_service, options);
InputStreamImpl is(&fs, &blocks, std::make_shared<BadDataNodeTracker>());
Status stat;
size_t read = 0;
struct Trait {
static void InitializeMockReader(MockReader *reader) {
EXPECT_CALL(*reader, async_connect(_, _, _, _, _, _))
.WillOnce(InvokeArgument<5>(Status::OK()));
EXPECT_CALL(*reader, async_read_some(_, _))
.Times(4)
.WillRepeatedly(InvokeArgument<1>(Status::OK(), sizeof(buf) / 4));
}
};
is.AsyncReadBlock<MockBlockReaderTrait<Trait>>(
"client", block, dn, 0, asio::buffer(buf, sizeof(buf)),
[&stat, &read](const Status &status, const std::string &,
size_t transferred) {
stat = status;
read = transferred;
});
ASSERT_TRUE(stat.ok());
ASSERT_EQ(sizeof(buf), read);
read = 0;
}
TEST(InputStreamTest, TestReadError) {
LocatedBlocksProto blocks;
LocatedBlockProto block;
DatanodeInfoProto dn;
char buf[4096] = {
0,
};
IoServiceImpl io_service;
Options options;
FileSystemImpl fs(&io_service, options);
InputStreamImpl is(&fs, &blocks, std::make_shared<BadDataNodeTracker>());
Status stat;
size_t read = 0;
struct Trait {
static void InitializeMockReader(MockReader *reader) {
EXPECT_CALL(*reader, async_connect(_, _, _, _, _, _))
.WillOnce(InvokeArgument<5>(Status::OK()));
EXPECT_CALL(*reader, async_read_some(_, _))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf) / 4))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf) / 4))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf) / 4))
.WillOnce(InvokeArgument<1>(Status::Error("error"), 0));
}
};
is.AsyncReadBlock<MockBlockReaderTrait<Trait>>(
"client", block, dn, 0, asio::buffer(buf, sizeof(buf)),
[&stat, &read](const Status &status, const std::string &,
size_t transferred) {
stat = status;
read = transferred;
});
ASSERT_FALSE(stat.ok());
ASSERT_EQ(sizeof(buf) / 4 * 3, read);
read = 0;
}
TEST(InputStreamTest, TestExcludeDataNode) {
LocatedBlocksProto blocks;
LocatedBlockProto *block = blocks.add_blocks();
ExtendedBlockProto *b = block->mutable_b();
b->set_poolid("");
b->set_blockid(1);
b->set_generationstamp(1);
b->set_numbytes(4096);
DatanodeInfoProto *di = block->add_locs();
DatanodeIDProto *dnid = di->mutable_id();
dnid->set_datanodeuuid("foo");
char buf[4096] = {
0,
};
IoServiceImpl io_service;
Options options;
FileSystemImpl fs(&io_service, options);
InputStreamImpl is(&fs, &blocks, std::make_shared<BadDataNodeTracker>());
Status stat;
size_t read = 0;
struct Trait {
static void InitializeMockReader(MockReader *reader) {
EXPECT_CALL(*reader, async_connect(_, _, _, _, _, _))
.WillOnce(InvokeArgument<5>(Status::OK()));
EXPECT_CALL(*reader, async_read_some(_, _))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf)));
}
};
std::shared_ptr<NodeExclusionRule> exclude_set =
std::make_shared<ExclusionSet>(std::set<std::string>({"foo"}));
is.AsyncPreadSome(0, asio::buffer(buf, sizeof(buf)), exclude_set,
[&stat, &read](const Status &status, const std::string &,
size_t transferred) {
stat = status;
read = transferred;
});
ASSERT_EQ(static_cast<int>(std::errc::resource_unavailable_try_again),
stat.code());
ASSERT_EQ(0UL, read);
}
int main(int argc, char *argv[]) {
// The following line must be executed to initialize Google Mock
// (and Google Test) before running the tests.
::testing::InitGoogleMock(&argc, argv);
return RUN_ALL_TESTS();
}

View File

@ -18,6 +18,8 @@
#ifndef LIBHDFSPP_TEST_MOCK_CONNECTION_H_
#define LIBHDFSPP_TEST_MOCK_CONNECTION_H_
#include "common/async_stream.h"
#include <asio/error_code.hpp>
#include <asio/buffer.hpp>
#include <asio/streambuf.hpp>
@ -27,13 +29,15 @@
namespace hdfs {
class MockConnectionBase {
class MockConnectionBase : public AsyncStream{
public:
MockConnectionBase(::asio::io_service *io_service);
virtual ~MockConnectionBase();
typedef std::pair<asio::error_code, std::string> ProducerResult;
template <class MutableBufferSequence, class Handler>
void async_read_some(const MutableBufferSequence &buf, Handler &&handler) {
void async_read_some(const MutableBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
if (produced_.size() == 0) {
ProducerResult r = Produce();
if (r.first) {
@ -51,8 +55,9 @@ public:
io_service_->post(std::bind(handler, asio::error_code(), len));
}
template <class ConstBufferSequence, class Handler>
void async_write_some(const ConstBufferSequence &buf, Handler &&handler) {
void async_write_some(const ConstBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
// CompletionResult res = OnWrite(buf);
io_service_->post(std::bind(handler, asio::error_code(), asio::buffer_size(buf)));
}

View File

@ -21,6 +21,8 @@
#include "datatransfer.pb.h"
#include "common/util.h"
#include "reader/block_reader.h"
#include "reader/datatransfer.h"
#include "reader/fileinfo.h"
#include <google/protobuf/io/coded_stream.h>
#include <google/protobuf/io/zero_copy_stream_impl.h>
@ -36,10 +38,14 @@ using ::hadoop::hdfs::DataTransferEncryptorMessageProto;
using ::hadoop::hdfs::ExtendedBlockProto;
using ::hadoop::hdfs::PacketHeaderProto;
using ::hadoop::hdfs::ReadOpChecksumInfoProto;
using ::hadoop::hdfs::LocatedBlockProto;
using ::hadoop::hdfs::LocatedBlocksProto;
using ::asio::buffer;
using ::asio::error_code;
using ::asio::mutable_buffers_1;
using ::testing::_;
using ::testing::InvokeArgument;
using ::testing::Return;
using std::make_pair;
using std::string;
@ -49,12 +55,47 @@ namespace pbio = pb::io;
namespace hdfs {
class MockDNConnection : public MockConnectionBase {
public:
class MockDNConnection : public MockConnectionBase, public DataNodeConnection{
public:
MockDNConnection(::asio::io_service &io_service)
: MockConnectionBase(&io_service) {}
MOCK_METHOD0(Produce, ProducerResult());
MOCK_METHOD1(Connect, void(std::function<void(Status status, std::shared_ptr<DataNodeConnection> dn)>));
void async_read_some(const MutableBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
this->MockConnectionBase::async_read_some(buf, handler);
}
void async_write_some(const ConstBuffers &buf,
std::function<void (const asio::error_code & error,
std::size_t bytes_transferred) > handler) override {
this->MockConnectionBase::async_write_some(buf, handler);
}
};
// Mocks AsyncReadPacket and AsyncRequestBlock but not AsyncReadBlock, so we
// can test the logic of AsyncReadBlock
class PartialMockReader : public BlockReaderImpl {
public:
PartialMockReader() :
BlockReaderImpl(BlockReaderOptions(), std::shared_ptr<DataNodeConnection>()) {};
MOCK_METHOD2(
AsyncReadPacket,
void(const asio::mutable_buffers_1 &,
const std::function<void(const Status &, size_t transferred)> &));
MOCK_METHOD5(AsyncRequestBlock,
void(const std::string &client_name,
const hadoop::hdfs::ExtendedBlockProto *block,
uint64_t length, uint64_t offset,
const std::function<void(Status)> &handler));
};
}
static inline string ToDelimitedString(const pb::MessageLite *msg) {
@ -94,20 +135,102 @@ static inline std::pair<error_code, string> ProducePacket(
return std::make_pair(error_code(), std::move(payload));
}
TEST(RemoteBlockReaderTest, TestReadSingleTrunk) {
auto file_info = std::make_shared<struct FileInfo>();
LocatedBlocksProto blocks;
LocatedBlockProto block;
char buf[4096] = {
0,
};
Status stat;
size_t read = 0;
PartialMockReader reader;
EXPECT_CALL(reader, AsyncRequestBlock(_, _, _, _, _))
.WillOnce(InvokeArgument<4>(Status::OK()));
EXPECT_CALL(reader, AsyncReadPacket(_, _))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf)));
reader.AsyncReadBlock(
GetRandomClientName(), block, 0, asio::buffer(buf, sizeof(buf)),
[&stat, &read](const Status &status, size_t transferred) {
stat = status;
read = transferred;
});
ASSERT_TRUE(stat.ok());
ASSERT_EQ(sizeof(buf), read);
read = 0;
}
TEST(RemoteBlockReaderTest, TestReadMultipleTrunk) {
LocatedBlockProto block;
char buf[4096] = {
0,
};
Status stat;
size_t read = 0;
PartialMockReader reader;
EXPECT_CALL(reader, AsyncRequestBlock(_, _, _, _, _))
.WillOnce(InvokeArgument<4>(Status::OK()));
EXPECT_CALL(reader, AsyncReadPacket(_, _))
.Times(4)
.WillRepeatedly(InvokeArgument<1>(Status::OK(), sizeof(buf) / 4));
reader.AsyncReadBlock(
GetRandomClientName(), block, 0, asio::buffer(buf, sizeof(buf)),
[&stat, &read](const Status &status, size_t transferred) {
stat = status;
read = transferred;
});
ASSERT_TRUE(stat.ok());
ASSERT_EQ(sizeof(buf), read);
read = 0;
}
TEST(RemoteBlockReaderTest, TestReadError) {
LocatedBlockProto block;
char buf[4096] = {
0,
};
Status stat;
size_t read = 0;
PartialMockReader reader;
EXPECT_CALL(reader, AsyncRequestBlock(_, _, _, _, _))
.WillOnce(InvokeArgument<4>(Status::OK()));
EXPECT_CALL(reader, AsyncReadPacket(_, _))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf) / 4))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf) / 4))
.WillOnce(InvokeArgument<1>(Status::OK(), sizeof(buf) / 4))
.WillOnce(InvokeArgument<1>(Status::Error("error"), 0));
reader.AsyncReadBlock(
GetRandomClientName(), block, 0, asio::buffer(buf, sizeof(buf)),
[&stat, &read](const Status &status, size_t transferred) {
stat = status;
read = transferred;
});
ASSERT_FALSE(stat.ok());
ASSERT_EQ(sizeof(buf) / 4 * 3, read);
read = 0;
}
template <class Stream = MockDNConnection, class Handler>
static std::shared_ptr<RemoteBlockReader<Stream>> ReadContent(
Stream *conn, TokenProto *token, const ExtendedBlockProto &block,
static std::shared_ptr<BlockReaderImpl>
ReadContent(std::shared_ptr<Stream> conn, const ExtendedBlockProto &block,
uint64_t length, uint64_t offset, const mutable_buffers_1 &buf,
const Handler &handler) {
BlockReaderOptions options;
auto reader = std::make_shared<RemoteBlockReader<Stream>>(options, conn);
auto reader = std::make_shared<BlockReaderImpl>(options, conn);
Status result;
reader->async_connect("libhdfs++", token, &block, length, offset,
reader->AsyncRequestBlock("libhdfs++", &block, length, offset,
[buf, reader, handler](const Status &stat) {
if (!stat.ok()) {
handler(stat, 0);
} else {
reader->async_read_some(buf, handler);
reader->AsyncReadPacket(buf, handler);
}
});
return reader;
@ -117,11 +240,11 @@ TEST(RemoteBlockReaderTest, TestReadWholeBlock) {
static const size_t kChunkSize = 512;
static const string kChunkData(kChunkSize, 'a');
::asio::io_service io_service;
MockDNConnection conn(io_service);
auto conn = std::make_shared<MockDNConnection>(io_service);
BlockOpResponseProto block_op_resp;
block_op_resp.set_status(::hadoop::hdfs::Status::SUCCESS);
EXPECT_CALL(conn, Produce())
EXPECT_CALL(*conn, Produce())
.WillOnce(Return(Produce(ToDelimitedString(&block_op_resp))))
.WillOnce(Return(ProducePacket(kChunkData, "", 0, 1, true)));
@ -130,16 +253,19 @@ TEST(RemoteBlockReaderTest, TestReadWholeBlock) {
block.set_blockid(0);
block.set_generationstamp(0);
bool done = false;
std::string data(kChunkSize, 0);
ReadContent(&conn, nullptr, block, kChunkSize, 0,
ReadContent(conn, block, kChunkSize, 0,
buffer(const_cast<char *>(data.c_str()), data.size()),
[&data, &io_service](const Status &stat, size_t transferred) {
[&data, &io_service, &done](const Status &stat, size_t transferred) {
ASSERT_TRUE(stat.ok());
ASSERT_EQ(kChunkSize, transferred);
ASSERT_EQ(kChunkData, data);
done = true;
io_service.stop();
});
io_service.run();
ASSERT_TRUE(done);
}
TEST(RemoteBlockReaderTest, TestReadWithinChunk) {
@ -149,7 +275,7 @@ TEST(RemoteBlockReaderTest, TestReadWithinChunk) {
static const string kChunkData = string(kOffset, 'a') + string(kLength, 'b');
::asio::io_service io_service;
MockDNConnection conn(io_service);
auto conn = std::make_shared<MockDNConnection>(io_service);
BlockOpResponseProto block_op_resp;
ReadOpChecksumInfoProto *checksum_info =
block_op_resp.mutable_readopchecksuminfo();
@ -159,7 +285,7 @@ TEST(RemoteBlockReaderTest, TestReadWithinChunk) {
checksum->set_bytesperchecksum(512);
block_op_resp.set_status(::hadoop::hdfs::Status::SUCCESS);
EXPECT_CALL(conn, Produce())
EXPECT_CALL(*conn, Produce())
.WillOnce(Return(Produce(ToDelimitedString(&block_op_resp))))
.WillOnce(Return(ProducePacket(kChunkData, "", kOffset, 1, true)));
@ -168,16 +294,20 @@ TEST(RemoteBlockReaderTest, TestReadWithinChunk) {
block.set_blockid(0);
block.set_generationstamp(0);
bool done = false;
string data(kLength, 0);
ReadContent(&conn, nullptr, block, data.size(), kOffset,
ReadContent(conn, block, data.size(), kOffset,
buffer(const_cast<char *>(data.c_str()), data.size()),
[&data, &io_service](const Status &stat, size_t transferred) {
[&data, &io_service,&done](const Status &stat, size_t transferred) {
ASSERT_TRUE(stat.ok());
ASSERT_EQ(kLength, transferred);
ASSERT_EQ(kChunkData.substr(kOffset, kLength), data);
done = true;
io_service.stop();
});
io_service.run();
ASSERT_TRUE(done);
}
TEST(RemoteBlockReaderTest, TestReadMultiplePacket) {
@ -185,11 +315,11 @@ TEST(RemoteBlockReaderTest, TestReadMultiplePacket) {
static const string kChunkData(kChunkSize, 'a');
::asio::io_service io_service;
MockDNConnection conn(io_service);
auto conn = std::make_shared<MockDNConnection>(io_service);
BlockOpResponseProto block_op_resp;
block_op_resp.set_status(::hadoop::hdfs::Status::SUCCESS);
EXPECT_CALL(conn, Produce())
EXPECT_CALL(*conn, Produce())
.WillOnce(Return(Produce(ToDelimitedString(&block_op_resp))))
.WillOnce(Return(ProducePacket(kChunkData, "", 0, 1, false)))
.WillOnce(Return(ProducePacket(kChunkData, "", kChunkSize, 2, true)));
@ -202,25 +332,22 @@ TEST(RemoteBlockReaderTest, TestReadMultiplePacket) {
string data(kChunkSize, 0);
mutable_buffers_1 buf = buffer(const_cast<char *>(data.c_str()), data.size());
BlockReaderOptions options;
auto reader =
std::make_shared<RemoteBlockReader<MockDNConnection>>(options, &conn);
auto reader = std::make_shared<BlockReaderImpl>(options, conn);
Status result;
reader->async_connect(
"libhdfs++", nullptr, &block, data.size(), 0,
reader->AsyncRequestBlock(
"libhdfs++", &block, data.size(), 0,
[buf, reader, &data, &io_service](const Status &stat) {
ASSERT_TRUE(stat.ok());
reader->async_read_some(
buf, [buf, reader, &data, &io_service](const Status &stat,
size_t transferred) {
reader->AsyncReadPacket(
buf, [buf, reader, &data, &io_service](const Status &stat, size_t transferred) {
ASSERT_TRUE(stat.ok());
ASSERT_EQ(kChunkSize, transferred);
ASSERT_EQ(kChunkData, data);
data.clear();
data.resize(kChunkSize);
transferred = 0;
reader->async_read_some(
buf,
[&data, &io_service](const Status &stat, size_t transferred) {
reader->AsyncReadPacket(
buf, [&data,&io_service](const Status &stat, size_t transferred) {
ASSERT_TRUE(stat.ok());
ASSERT_EQ(kChunkSize, transferred);
ASSERT_EQ(kChunkData, data);
@ -234,12 +361,11 @@ TEST(RemoteBlockReaderTest, TestReadMultiplePacket) {
TEST(RemoteBlockReaderTest, TestSaslConnection) {
static const size_t kChunkSize = 512;
static const string kChunkData(kChunkSize, 'a');
static const string kAuthPayload =
"realm=\"0\",nonce=\"+GAWc+O6yEAWpew/"
static const string kAuthPayload = "realm=\"0\",nonce=\"+GAWc+O6yEAWpew/"
"qKah8qh4QZLoOLCDcTtEKhlS\",qop=\"auth\","
"charset=utf-8,algorithm=md5-sess";
::asio::io_service io_service;
MockDNConnection conn(io_service);
auto conn = std::make_shared<MockDNConnection>(io_service);
BlockOpResponseProto block_op_resp;
block_op_resp.set_status(::hadoop::hdfs::Status::SUCCESS);
@ -252,23 +378,23 @@ TEST(RemoteBlockReaderTest, TestSaslConnection) {
::hadoop::hdfs::
DataTransferEncryptorMessageProto_DataTransferEncryptorStatus_SUCCESS);
EXPECT_CALL(conn, Produce())
EXPECT_CALL(*conn, Produce())
.WillOnce(Return(Produce(ToDelimitedString(&sasl_resp0))))
.WillOnce(Return(Produce(ToDelimitedString(&sasl_resp1))))
.WillOnce(Return(Produce(ToDelimitedString(&block_op_resp))))
.WillOnce(Return(ProducePacket(kChunkData, "", 0, 1, true)));
DataTransferSaslStream<MockDNConnection> sasl_conn(&conn, "foo", "bar");
auto sasl_conn = std::make_shared<DataTransferSaslStream<MockDNConnection> >(conn, "foo", "bar");
ExtendedBlockProto block;
block.set_poolid("foo");
block.set_blockid(0);
block.set_generationstamp(0);
std::string data(kChunkSize, 0);
sasl_conn.Handshake([&sasl_conn, &block, &data, &io_service](
sasl_conn->Handshake([sasl_conn, &block, &data, &io_service](
const Status &s) {
ASSERT_TRUE(s.ok());
ReadContent(&sasl_conn, nullptr, block, kChunkSize, 0,
ReadContent(sasl_conn, block, kChunkSize, 0,
buffer(const_cast<char *>(data.c_str()), data.size()),
[&data, &io_service](const Status &stat, size_t transferred) {
ASSERT_TRUE(stat.ok());