add prophet-rocksdb

.gitignore

@@ -1,6 +1,9 @@
 make_config.mk
 rocksdb.pc
 
+out.txt
+
+
 *.a
 *.arc
 *.d

Makefile

@@ -10,6 +10,7 @@ BASH_EXISTS := $(shell which bash)
 SHELL := $(shell which bash)
 include common.mk
 
+USE_RTTI = 1
 CLEAN_FILES = # deliberately empty, so we can append below.
 CFLAGS += ${EXTRA_CFLAGS}
 CXXFLAGS += ${EXTRA_CXXFLAGS}

README.md

@@ -1,31 +1,38 @@
-## RocksDB: A Persistent Key-Value Store for Flash and RAM Storage
+## Prophet
 
-[![CircleCI Status](https://circleci.com/gh/facebook/rocksdb.svg?style=svg)](https://circleci.com/gh/facebook/rocksdb)
-[![Appveyor Build status](https://ci.appveyor.com/api/projects/status/fbgfu0so3afcno78/branch/main?svg=true)](https://ci.appveyor.com/project/Facebook/rocksdb/branch/main)
-[![PPC64le Build Status](http://140-211-168-68-openstack.osuosl.org:8080/buildStatus/icon?job=rocksdb&style=plastic)](http://140-211-168-68-openstack.osuosl.org:8080/job/rocksdb)
+Build Prophet:
 
-RocksDB is developed and maintained by Facebook Database Engineering Team.
-It is built on earlier work on [LevelDB](https://github.com/google/leveldb) by Sanjay Ghemawat (sanjay@google.com)
-and Jeff Dean (jeff@google.com)
+Please make sure you have installed the required dependencies in [RocksDB](https://github.com/facebook/rocksdb/blob/main/INSTALL.md) and replace `<zoned block device>` to real ZNS SSD device name.
 
-This code is a library that forms the core building block for a fast
-key-value server, especially suited for storing data on flash drives.
-It has a Log-Structured-Merge-Database (LSM) design with flexible tradeoffs
-between Write-Amplification-Factor (WAF), Read-Amplification-Factor (RAF)
-and Space-Amplification-Factor (SAF). It has multi-threaded compactions,
-making it especially suitable for storing multiple terabytes of data in a
-single database.
+```bash
+sudo git clone https://github.com/Flappybird11101001/prophet-rocksdb.git rocksdb
+cd rocksdb
+sudo git clone https://github.com/Flappybird11101001/prophet-zenfs.git plugin/zenfs
+sudo DISABLE_WARNING_AS_ERROR=1 ROCKSDB_PLUGINS=zenfs make -j db_bench install DEBUG_LEVEL=0
+pushd .
+cd plugin/zenfs/util
+sudo make
+popd
+```
 
-Start with example usage here: https://github.com/facebook/rocksdb/tree/main/examples
+initialize ZNS SSD device
 
-See the [github wiki](https://github.com/facebook/rocksdb/wiki) for more explanation.
+```bash
+echo deadline > /sys/class/block/<zoned block device>/queue/scheduler
+sudo ./plugin/zenfs/util/zenfs mkfs --zbd=<zoned block device> --aux_path=./temp --force
+```
 
-The public interface is in `include/`.  Callers should not include or
-rely on the details of any other header files in this package.  Those
-internal APIs may be changed without warning.
+# Benchmark
 
-Questions and discussions are welcome on the [RocksDB Developers Public](https://www.facebook.com/groups/rocksdb.dev/) Facebook group and [email list](https://groups.google.com/g/rocksdb) on Google Groups.
+run db_bench to test(the same config with paper in 64MB SST file size). 
 
-## License
+```bash
+sudo ./db_bench -num=400000000 -key_size=8 -value_size=256  -statistics=true -max_bytes_for_level_base=268435456  -target_file_size_base=67108864 -write_buffer_size=134217728 writable_file_max_buffer_size=134217728  -max_bytes_for_level_multiplier=4 -max_background_compactions=1 -max_background_flushes=1 -max_background_jobs=1 -soft_pending_compaction_bytes_limit=67108864 -hard_pending_compaction_bytes_limit=67108864 -level0_stop_writes_trigger=12 -level0_slowdown_writes_trigger=8 -level0_file_num_compaction_trigger=4 -max_write_buffer_number=1  -threads=1 -compaction_pri=4 -open_files=1000 -target_file_size_multiplier=1 --fs_uri=zenfs://dev:<zoned block device> --benchmarks='fillrandom,stats' --use_direct_io_for_flush_and_compaction
+```
 
-RocksDB is dual-licensed under both the GPLv2 (found in the COPYING file in the root directory) and Apache 2.0 License (found in the LICENSE.Apache file in the root directory).  You may select, at your option, one of the above-listed licenses.
+
+![allocation_migrated_data](./allocation_migrated_data.jpg)
+
+![allocation_wa](./allocation_wa.jpg)
+
+![allocation_zone_number_page-0001](./allocation_zone_number.jpg)

clear.sh

@@ -0,0 +1,7 @@
+rm -f level.out 
+rm -f lifetime.out 
+rm -f number_life.out 
+rm -f factor.out 
+rm -f last_compact.out
+rm -f rank.out 
+rm -rf clock.out

clock_pic.py

@@ -0,0 +1,31 @@
+import matplotlib.pyplot as plt
+import numpy as np
+
+prev_list = []
+tmp_prev_flush_list = []
+prev_flush_list = []
+type_list = []
+tot = 0
+for line in open("clock.out"): 
+    tot = tot + 1
+    if(tot != 1):
+        prev_list.append(int(line.split(' ')[0]))
+        tmp_prev_flush_list.append(int(line.split(' ')[1])) 
+        type_list.append(int(line.split(' ')[2]))
+
+
+y = np.array(prev_list)
+plt.hist(prev_list, bins=100, color="brown")
+plt.show()
+
+
+tot = 0
+for i in range(0, len(type_list)):
+    tot = tot + 1
+    if i + 1 < len(type_list) and type_list[i] == 2 and type_list[i + 1] == 1:
+        prev_flush_list.append(tmp_prev_flush_list[i])
+
+
+# y = np.array(prev_flush_list)
+plt.hist(prev_flush_list, bins=100, color="brown")
+plt.show()

db/builder.cc

@@ -43,6 +43,12 @@
 
 namespace ROCKSDB_NAMESPACE {
 
+extern void get_predict(int level, const FileMetaData &file, Version *v, const Compaction* compaction_, int &predict_, int &predict_type_, int &tmp_rank);
+extern void set_deleted_time(int fnumber, int clock);
+extern void update_fname(uint64_t id, std::string name);
+extern std::string get_fname(uint64_t id);
+extern int get_clock();
+
 class TableFactory;
 
 TableBuilder* NewTableBuilder(const TableBuilderOptions& tboptions,
@@ -147,10 +153,31 @@ Status BuildTable(
       bool use_direct_writes = file_options.use_direct_writes;
       TEST_SYNC_POINT_CALLBACK("BuildTable:create_file", &use_direct_writes);
 #endif  // !NDEBUG
-      IOStatus io_s = NewWritableFile(fs, fname, &file, file_options);
+      //file_options.lifetime = 1000;
+      FileOptions tmp_file_options = file_options;
+      tmp_file_options.lifetime = 100;
+
+      update_fname(meta->fd.GetNumber(), fname);
+      //在这里写入
+      IOStatus io_s = NewWritableFile(fs, fname, &file, tmp_file_options);
+
+
+      int predict;
+      int predict_type;
+      int rank;
+      const int output_level = 0;
+
+      get_predict(output_level, *meta, versions->GetColumnFamilySet()->GetDefault()->current(), nullptr, predict, predict_type, rank);
+      set_deleted_time(meta->fnumber, predict + get_clock());
+      printf("meta->fname=%s get_clock=%d lifetime=%d\n", fname.c_str(), get_clock(), predict + get_clock());
+      fs->SetFileLifetime(fname, predict + get_clock(), get_clock(), 0, output_level, std::vector<std::string> {});
+
+
+
+
       assert(s.ok());
       s = io_s;
-      if (io_status->ok()) {
+      if (io_status->ok()) {  
         *io_status = io_s;
       }
       if (!s.ok()) {

db/column_family.cc

@@ -1118,7 +1118,7 @@ Compaction* ColumnFamilyData::PickCompaction(
                imm_.current()->GetEarliestSequenceNumber(false));
   auto* result = compaction_picker_->PickCompaction(
       GetName(), mutable_options, mutable_db_options, current_->storage_info(),
-      log_buffer, earliest_mem_seqno);
+      log_buffer, earliest_mem_seqno); //PickCompaction来选择需要被Compact的文件
   if (result != nullptr) {
     result->SetInputVersion(current_);
   }

db/compaction/compaction.h

@@ -49,6 +49,7 @@ struct AtomicCompactionUnitBoundary {
   const InternalKey* largest = nullptr;
 };
 
+//使用此结构维护同一个level中所有的SST files
 // The structure that manages compaction input files associated
 // with the same physical level.
 struct CompactionInputFiles {
@@ -438,6 +439,7 @@ class Compaction {
   bool l0_files_might_overlap_;
 
   // Compaction input files organized by level. Constant after construction
+  //Compaction中的输入变量
   const std::vector<CompactionInputFiles> inputs_;
 
   // A copy of inputs_, organized more closely in memory