options: add recycle_log_file_num option

Signed-off-by: Sage Weil <sage@redhat.com>

db/c.cc

@@ -1687,6 +1687,11 @@ void rocksdb_options_set_keep_log_file_num(rocksdb_options_t* opt, size_t v) {
   opt->rep.keep_log_file_num = v;
 }
 
+void rocksdb_options_set_recycle_log_file_num(rocksdb_options_t* opt,
+                                              size_t v) {
+  opt->rep.recycle_log_file_num = v;
+}
+
 void rocksdb_options_set_soft_rate_limit(rocksdb_options_t* opt, double v) {
   opt->rep.soft_rate_limit = v;
 }

include/rocksdb/c.h

@@ -572,6 +572,8 @@ extern ROCKSDB_LIBRARY_API void rocksdb_options_set_log_file_time_to_roll(
     rocksdb_options_t*, size_t);
 extern ROCKSDB_LIBRARY_API void rocksdb_options_set_keep_log_file_num(
     rocksdb_options_t*, size_t);
+extern ROCKSDB_LIBRARY_API void rocksdb_options_set_recycle_log_file_num(
+    rocksdb_options_t*, size_t);
 extern ROCKSDB_LIBRARY_API void rocksdb_options_set_soft_rate_limit(
     rocksdb_options_t*, double);
 extern ROCKSDB_LIBRARY_API void rocksdb_options_set_hard_rate_limit(

include/rocksdb/options.h

@@ -957,6 +957,16 @@ struct DBOptions {
   // Default: 1000
   size_t keep_log_file_num;
 
+  // Recycle log files.
+  // If non-zero, we will reuse previously written log files for new
+  // logs, overwriting the old data.  The value indicates how many
+  // such files we will keep around at any point in time for later
+  // use.  This is more efficient because the blocks are already
+  // allocated and fdatasync does not need to update the inode after
+  // each write.
+  // Default: 0
+  size_t recycle_log_file_num;
+
   // manifest file is rolled over on reaching this limit.
   // The older manifest file be deleted.
   // The default value is MAX_INT so that roll-over does not take place.

java/rocksjni/options.cc

@@ -576,6 +576,33 @@ void Java_org_rocksdb_Options_setKeepLogFileNum(
   }
 }
 
+/*
+ * Class:     org_rocksdb_Options
+ * Method:    recycleLogFiles
+ * Signature: (J)J
+ */
+jlong Java_org_rocksdb_Options_recycleLogFileNum(JNIEnv* env, jobject jobj,
+                                                 jlong jhandle) {
+  return reinterpret_cast<rocksdb::Options*>(jhandle)->recycle_log_file_num;
+}
+
+/*
+ * Class:     org_rocksdb_Options
+ * Method:    setRecycleLogFiles
+ * Signature: (JJ)V
+ */
+void Java_org_rocksdb_Options_setRecycleLogFiles(JNIEnv* env, jobject jobj,
+                                                 jlong jhandle,
+                                                 jlong recycle_log_file_num) {
+  rocksdb::Status s = rocksdb::check_if_jlong_fits_size_t(recycle_log_file_num);
+  if (s.ok()) {
+    reinterpret_cast<rocksdb::Options*>(jhandle)->recycle_log_file_num =
+        recycle_log_file_num;
+  } else {
+    rocksdb::IllegalArgumentExceptionJni::ThrowNew(env, s);
+  }
+}
+
 /*
  * Class:     org_rocksdb_Options
  * Method:    maxManifestFileSize
@@ -3533,6 +3560,32 @@ jlong Java_org_rocksdb_DBOptions_keepLogFileNum(
   return reinterpret_cast<rocksdb::DBOptions*>(jhandle)->keep_log_file_num;
 }
 
+/*
+ * Class:     org_rocksdb_DBOptions
+ * Method:    setRecycleLogFiles
+ * Signature: (JJ)V
+ */
+void Java_org_rocksdb_DBOptions_setRecycleLogFileNum(
+    JNIEnv* env, jobject jobj, jlong jhandle, jlong recycle_log_file_num) {
+  rocksdb::Status s = rocksdb::check_if_jlong_fits_size_t(recycle_log_file_num);
+  if (s.ok()) {
+    reinterpret_cast<rocksdb::DBOptions*>(jhandle)->recycle_log_file_num =
+        recycle_log_file_num;
+  } else {
+    rocksdb::IllegalArgumentExceptionJni::ThrowNew(env, s);
+  }
+}
+
+/*
+ * Class:     org_rocksdb_DBOptions
+ * Method:    recycleLogFiles
+ * Signature: (J)J
+ */
+jlong Java_org_rocksdb_DBOptions_recycleLogFileNum(JNIEnv* env, jobject jobj,
+                                                   jlong jhandle) {
+  return reinterpret_cast<rocksdb::DBOptions*>(jhandle)->recycle_log_file_num;
+}
+
 /*
  * Class:     org_rocksdb_DBOptions
  * Method:    setMaxManifestFileSize

util/options.cc

@@ -231,6 +231,7 @@ DBOptions::DBOptions()
       max_log_file_size(0),
       log_file_time_to_roll(0),
       keep_log_file_num(1000),
+      recycle_log_file_num(0),
       max_manifest_file_size(std::numeric_limits<uint64_t>::max()),
       table_cache_numshardbits(4),
       WAL_ttl_seconds(0),
@@ -285,6 +286,7 @@ DBOptions::DBOptions(const Options& options)
       max_log_file_size(options.max_log_file_size),
       log_file_time_to_roll(options.log_file_time_to_roll),
       keep_log_file_num(options.keep_log_file_num),
+      recycle_log_file_num(options.recycle_log_file_num),
       max_manifest_file_size(options.max_manifest_file_size),
       table_cache_numshardbits(options.table_cache_numshardbits),
       WAL_ttl_seconds(options.WAL_ttl_seconds),
@@ -338,6 +340,8 @@ void DBOptions::Dump(Logger* log) const {
          log_file_time_to_roll);
     Header(log, "     Options.keep_log_file_num: %" ROCKSDB_PRIszt,
          keep_log_file_num);
+    Header(log, "  Options.recycle_log_file_num: %" ROCKSDB_PRIszt,
+           recycle_log_file_num);
     Header(log, "       Options.allow_os_buffer: %d", allow_os_buffer);
     Header(log, "      Options.allow_mmap_reads: %d", allow_mmap_reads);
     Header(log, "      Options.allow_fallocate: %d", allow_fallocate);

util/options_helper.h

@@ -207,6 +207,9 @@ static std::unordered_map<std::string, OptionTypeInfo> db_options_type_info = {
     {"keep_log_file_num",
      {offsetof(struct DBOptions, keep_log_file_num), OptionType::kSizeT,
       OptionVerificationType::kNormal}},
+    {"recycle_log_file_num",
+     {offsetof(struct DBOptions, recycle_log_file_num), OptionType::kSizeT,
+      OptionVerificationType::kNormal}},
     {"log_file_time_to_roll",
      {offsetof(struct DBOptions, log_file_time_to_roll), OptionType::kSizeT,
       OptionVerificationType::kNormal}},

util/options_test.cc

@@ -323,6 +323,7 @@ TEST_F(OptionsTest, GetOptionsFromMapTest) {
       {"max_log_file_size", "37"},
       {"log_file_time_to_roll", "38"},
       {"keep_log_file_num", "39"},
+      {"recycle_log_file_num", "5"},
       {"max_manifest_file_size", "40"},
       {"table_cache_numshardbits", "41"},
       {"WAL_ttl_seconds", "43"},
@@ -339,7 +340,8 @@ TEST_F(OptionsTest, GetOptionsFromMapTest) {
       {"new_table_reader_for_compaction_inputs", "true"},
       {"compaction_readahead_size", "100"},
       {"bytes_per_sync", "47"},
-      {"wal_bytes_per_sync", "48"}, };
+      {"wal_bytes_per_sync", "48"},
+  };
 
   ColumnFamilyOptions base_cf_opt;
   ColumnFamilyOptions new_cf_opt;
@@ -431,6 +433,7 @@ TEST_F(OptionsTest, GetOptionsFromMapTest) {
   ASSERT_EQ(new_db_opt.max_log_file_size, 37U);
   ASSERT_EQ(new_db_opt.log_file_time_to_roll, 38U);
   ASSERT_EQ(new_db_opt.keep_log_file_num, 39U);
+  ASSERT_EQ(new_db_opt.recycle_log_file_num, 5U);
   ASSERT_EQ(new_db_opt.max_manifest_file_size, static_cast<uint64_t>(40));
   ASSERT_EQ(new_db_opt.table_cache_numshardbits, 41);
   ASSERT_EQ(new_db_opt.WAL_ttl_seconds, static_cast<uint64_t>(43));
@@ -692,6 +695,7 @@ void RandomInitDBOptions(DBOptions* db_opt, Random* rnd) {
   db_opt->skip_stats_update_on_db_open = rnd->Uniform(2);
   db_opt->use_adaptive_mutex = rnd->Uniform(2);
   db_opt->use_fsync = rnd->Uniform(2);
+  db_opt->recycle_log_file_num = rnd->Uniform(2);
 
   // int options
   db_opt->max_background_compactions = rnd->Uniform(100);