1 /** 2 * Copyright 2020-2022 Huawei Technologies Co., Ltd 3 * 4 * Licensed under the Apache License, Version 2.0 (the "License"); 5 * you may not use this file except in compliance with the License. 6 * You may obtain a copy of the License at 7 * 8 * http://www.apache.org/licenses/LICENSE-2.0 9 * 10 * Unless required by applicable law or agreed to in writing, software 11 * distributed under the License is distributed on an "AS IS" BASIS, 12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 13 * See the License for the specific language governing permissions and 14 * limitations under the License. 15 */ 16 #ifndef MINDSPORE_CCSRC_MINDDATA_DATASET_CONNECTOR_SIZE_H 17 #define MINDSPORE_CCSRC_MINDDATA_DATASET_CONNECTOR_SIZE_H 18 19 #include <string> 20 #include <vector> 21 #include <nlohmann/json.hpp> 22 #include "minddata/dataset/engine/perf/profiling.h" 23 #include "minddata/dataset/engine/datasetops/dataset_op.h" 24 25 using json = nlohmann::json; 26 27 namespace mindspore { 28 namespace dataset { 29 class ExecutionTree; 30 31 // Connector size sampling samples the output connector size of each op in the pipeline. 32 // It support JSON serialization for external usage. 33 class ConnectorSize : public Sampling { 34 // Connector size sampling data is stored as a 2D vector 35 // op_0 ... op_m 36 // sample_0 size_0_0 ... size_m_0 37 // ... ... ... ... 38 // sample_n size_0_m ... size_m_n 39 // 40 // A circular buffer will be implemented in the future to make this table more flexible. 41 using ConnectorSizeSample = std::vector<int>; 42 using ConnectorSizeSampleTable = std::vector<ConnectorSizeSample>; 43 using Timestamps = std::vector<uint64_t>; 44 45 public: ConnectorSize(ExecutionTree * tree)46 explicit ConnectorSize(ExecutionTree *tree) : tree_(tree) {} 47 48 ~ConnectorSize() override = default; 49 50 // Driver function for connector size sampling. 51 // This function samples the connector size of every nodes within the ExecutionTree 52 Status Sample() override; 53 Name()54 std::string Name() const override { return kConnectorSizeSamplingName; } 55 56 // Save sampling data to file 57 // @return Status The status code returned 58 Status SaveToFile(const std::string &dir_path, const std::string &rank_id) override; 59 60 Status Init() override; 61 62 // Parse op information and transform to json format 63 json ParseOpInfo(const DatasetOp &node) const; 64 65 // Change file mode after save throughput data ChangeFileMode(const std::string & dir_path,const std::string & rank_id)66 Status ChangeFileMode(const std::string &dir_path, const std::string &rank_id) override { return Status::OK(); } 67 68 // Get the vector of connector sizes of given op for samples taken between start and end time 69 Status GetOpConnectorSize(int32_t op_id, uint64_t start_time, uint64_t end_time, std::vector<int32_t> *result); 70 71 // Clear all collected data 72 void Clear() override; 73 74 protected: 75 Path GetFileName(const std::string &dir_path, const std::string &rank_id) override; 76 77 private: 78 json initial_nodes_data; // store data when execution tree is running. (all information for ops except sampled data) 79 ExecutionTree *tree_ = nullptr; // ExecutionTree pointer 80 ConnectorSizeSampleTable sample_table_; // Dataset structure to store all samples of connector size sampling 81 Timestamps ts_; // time of sample 82 }; 83 84 } // namespace dataset 85 } // namespace mindspore 86 87 #endif // MINDSPORE_CCSRC_MINDDATA_DATASET_CONNECTOR_SIZE_H 88