Cherry pick to v3.0.0 (0110-0114)

.github/workflows/pull_request.yml

@@ -170,6 +170,11 @@ jobs:
           make RM_DIR=false DEBUG=false J=${{ steps.cmake.outputs.j }} tck
         working-directory: tests/
         timeout-minutes: 60
+      - name: LDBC
+        run: |
+          make RM_DIR=false DEBUG=false J=${{ steps.cmake.outputs.j }} ldbc
+        working-directory: tests/
+        timeout-minutes: 60
       - name: Down cluster
         run: |
           make RM_DIR=false down

conf/nebula-graphd.conf.default

@@ -84,8 +84,9 @@
 # System memory high watermark ratio, cancel the memory checking when the ratio greater than 1.0
 --system_memory_high_watermark_ratio=0.8
 
+########## metrics ##########
+--enable_space_level_metrics=false
+
 ########## experimental feature ##########
 # if use experimental features
 --enable_experimental_feature=false
-
---enable_space_level_metrics=false

conf/nebula-graphd.conf.production

@@ -83,8 +83,9 @@
 # System memory high watermark ratio, cancel the memory checking when the ratio greater than 1.0
 --system_memory_high_watermark_ratio=0.8
 
+########## metrics ##########
+--enable_space_level_metrics=false
+
 ########## experimental feature ##########
 # if use experimental features
 --enable_experimental_feature=false
-
---enable_space_level_metrics=false

src/clients/meta/MetaClient.cpp

@@ -364,8 +364,8 @@ bool MetaClient::loadData() {
     GraphSpaceID spaceId = spaceInfo.first;
     std::shared_ptr<SpaceInfoCache> info = spaceInfo.second;
     std::shared_ptr<SpaceInfoCache> infoDeepCopy = std::make_shared<SpaceInfoCache>(*info);
-    infoDeepCopy->tagSchemas_ = buildTagSchemas(infoDeepCopy->tagItemVec_, &infoDeepCopy->pool_);
-    infoDeepCopy->edgeSchemas_ = buildEdgeSchemas(infoDeepCopy->edgeItemVec_, &infoDeepCopy->pool_);
+    infoDeepCopy->tagSchemas_ = buildTagSchemas(infoDeepCopy->tagItemVec_);
+    infoDeepCopy->edgeSchemas_ = buildEdgeSchemas(infoDeepCopy->edgeItemVec_);
     infoDeepCopy->tagIndexes_ = buildIndexes(infoDeepCopy->tagIndexItemVec_);
     infoDeepCopy->edgeIndexes_ = buildIndexes(infoDeepCopy->edgeIndexItemVec_);
     newMetaData->localCache_[spaceId] = infoDeepCopy;
@@ -396,14 +396,14 @@ bool MetaClient::loadData() {
   return true;
 }
 
-TagSchemas MetaClient::buildTagSchemas(std::vector<cpp2::TagItem> tagItemVec, ObjectPool* pool) {
+TagSchemas MetaClient::buildTagSchemas(std::vector<cpp2::TagItem> tagItemVec) {
   TagSchemas tagSchemas;
   TagID lastTagId = -1;
   for (auto& tagIt : tagItemVec) {
     // meta will return the different version from new to old
     auto schema = std::make_shared<NebulaSchemaProvider>(tagIt.get_version());
     for (const auto& colIt : tagIt.get_schema().get_columns()) {
-      addSchemaField(schema.get(), colIt, pool);
+      addSchemaField(schema.get(), colIt);
     }
     // handle schema property
     schema->setProp(tagIt.get_schema().get_schema_prop());
@@ -417,16 +417,15 @@ TagSchemas MetaClient::buildTagSchemas(std::vector<cpp2::TagItem> tagItemVec, Ob
   return tagSchemas;
 }
 
-EdgeSchemas MetaClient::buildEdgeSchemas(std::vector<cpp2::EdgeItem> edgeItemVec,
-                                         ObjectPool* pool) {
+EdgeSchemas MetaClient::buildEdgeSchemas(std::vector<cpp2::EdgeItem> edgeItemVec) {
   EdgeSchemas edgeSchemas;
   std::unordered_set<std::pair<GraphSpaceID, EdgeType>> edges;
   EdgeType lastEdgeType = -1;
   for (auto& edgeIt : edgeItemVec) {
     // meta will return the different version from new to old
     auto schema = std::make_shared<NebulaSchemaProvider>(edgeIt.get_version());
     for (const auto& col : edgeIt.get_schema().get_columns()) {
-      MetaClient::addSchemaField(schema.get(), col, pool);
+      MetaClient::addSchemaField(schema.get(), col);
     }
     // handle shcem property
     schema->setProp(edgeIt.get_schema().get_schema_prop());
@@ -440,32 +439,19 @@ EdgeSchemas MetaClient::buildEdgeSchemas(std::vector<cpp2::EdgeItem> edgeItemVec
   return edgeSchemas;
 }
 
-void MetaClient::addSchemaField(NebulaSchemaProvider* schema,
-                                const cpp2::ColumnDef& col,
-                                ObjectPool* pool) {
+void MetaClient::addSchemaField(NebulaSchemaProvider* schema, const cpp2::ColumnDef& col) {
   bool hasDef = col.default_value_ref().has_value();
   auto& colType = col.get_type();
   size_t len = colType.type_length_ref().has_value() ? *colType.get_type_length() : 0;
   cpp2::GeoShape geoShape =
       colType.geo_shape_ref().has_value() ? *colType.get_geo_shape() : cpp2::GeoShape::ANY;
   bool nullable = col.nullable_ref().has_value() ? *col.get_nullable() : false;
-  Expression* defaultValueExpr = nullptr;
+  std::string encoded;
   if (hasDef) {
-    auto encoded = *col.get_default_value();
-    defaultValueExpr = Expression::decode(pool, folly::StringPiece(encoded.data(), encoded.size()));
-
-    if (defaultValueExpr == nullptr) {
-      LOG(ERROR) << "Wrong expr default value for column name: " << col.get_name();
-      hasDef = false;
-    }
+    encoded = *col.get_default_value();
   }
 
-  schema->addField(col.get_name(),
-                   colType.get_type(),
-                   len,
-                   nullable,
-                   hasDef ? defaultValueExpr : nullptr,
-                   geoShape);
+  schema->addField(col.get_name(), colType.get_type(), len, nullable, encoded, geoShape);
 }
 
 bool MetaClient::loadSchemas(GraphSpaceID spaceId,
@@ -493,9 +479,9 @@ bool MetaClient::loadSchemas(GraphSpaceID spaceId,
   auto edgeItemVec = edgeRet.value();
   allEdgeMap[spaceId] = {};
   spaceInfoCache->tagItemVec_ = tagItemVec;
-  spaceInfoCache->tagSchemas_ = buildTagSchemas(tagItemVec, &spaceInfoCache->pool_);
+  spaceInfoCache->tagSchemas_ = buildTagSchemas(tagItemVec);
   spaceInfoCache->edgeItemVec_ = edgeItemVec;
-  spaceInfoCache->edgeSchemas_ = buildEdgeSchemas(edgeItemVec, &spaceInfoCache->pool_);
+  spaceInfoCache->edgeSchemas_ = buildEdgeSchemas(edgeItemVec);
 
   for (auto& tagIt : tagItemVec) {
     tagNameIdMap.emplace(std::make_pair(spaceId, tagIt.get_tag_name()), tagIt.get_tag_id());
@@ -859,6 +845,10 @@ Status MetaClient::handleResponse(const RESP& resp) {
       return Status::Error("Invalid param!");
     case nebula::cpp2::ErrorCode::E_WRONGCLUSTER:
       return Status::Error("Wrong cluster!");
+    case nebula::cpp2::ErrorCode::E_ZONE_NOT_ENOUGH:
+      return Status::Error("Zone not enough!");
+    case nebula::cpp2::ErrorCode::E_ZONE_IS_EMPTY:
+      return Status::Error("Zone is empty!");
     case nebula::cpp2::ErrorCode::E_STORE_FAILURE:
       return Status::Error("Store failure!");
     case nebula::cpp2::ErrorCode::E_STORE_SEGMENT_ILLEGAL:

src/clients/meta/MetaClient.h

@@ -91,8 +91,6 @@ struct SpaceInfoCache {
   std::vector<cpp2::IndexItem> edgeIndexItemVec_;
   Indexes edgeIndexes_;
   Listeners listeners_;
-  // objPool used to decode when adding field
-  ObjectPool pool_;
   std::unordered_map<PartitionID, TermID> termOfPartition_;
 
   SpaceInfoCache() = default;
@@ -816,10 +814,10 @@ class MetaClient {
     ServiceClientsList serviceClientList_;
   };
 
-  void addSchemaField(NebulaSchemaProvider* schema, const cpp2::ColumnDef& col, ObjectPool* pool);
+  void addSchemaField(NebulaSchemaProvider* schema, const cpp2::ColumnDef& col);
 
-  TagSchemas buildTagSchemas(std::vector<cpp2::TagItem> tagItemVec, ObjectPool* pool);
-  EdgeSchemas buildEdgeSchemas(std::vector<cpp2::EdgeItem> edgeItemVec, ObjectPool* pool);
+  TagSchemas buildTagSchemas(std::vector<cpp2::TagItem> tagItemVec);
+  EdgeSchemas buildEdgeSchemas(std::vector<cpp2::EdgeItem> edgeItemVec);
 
   std::unique_ptr<thread::GenericWorker> bgThread_;
   SpaceNameIdMap spaceIndexByName_;

src/codec/RowWriterV2.cpp

@@ -815,7 +815,9 @@ WriteResult RowWriterV2::checkUnsetFields() noexcept {
 
       WriteResult r = WriteResult::SUCCEEDED;
       if (field->hasDefault()) {
-        auto expr = field->defaultValue()->clone();
+        ObjectPool pool;
+        auto& exprStr = field->defaultValue();
+        auto expr = Expression::decode(&pool, folly::StringPiece(exprStr.data(), exprStr.size()));
         auto defVal = Expression::eval(expr, expCtx);
         switch (defVal.type()) {
           case Value::Type::NULLVALUE:
@@ -851,7 +853,7 @@ WriteResult RowWriterV2::checkUnsetFields() noexcept {
           default:
             LOG(FATAL) << "Unsupported default value type: " << defVal.typeName()
                        << ", default value: " << defVal
-                       << ", default value expr: " << field->defaultValue()->toString();
+                       << ", default value expr: " << field->defaultValue();
         }
       } else {
         // Set NULL

src/codec/test/ResultSchemaProvider.cpp

@@ -22,7 +22,7 @@ ResultSchemaProvider::ResultSchemaField::ResultSchemaField(std::string name,
                                                            bool nullable,
                                                            int32_t offset,
                                                            size_t nullFlagPos,
-                                                           Expression* defaultValue,
+                                                           std::string defaultValue,
                                                            meta::cpp2::GeoShape geoShape)
     : name_(std::move(name)),
       type_(type),
@@ -42,14 +42,14 @@ PropertyType ResultSchemaProvider::ResultSchemaField::type() const {
 }
 
 bool ResultSchemaProvider::ResultSchemaField::hasDefault() const {
-  return defaultValue_ != nullptr;
+  return defaultValue_ != "";
 }
 
 bool ResultSchemaProvider::ResultSchemaField::nullable() const {
   return nullable_;
 }
 
-Expression* ResultSchemaProvider::ResultSchemaField::defaultValue() const {
+const std::string& ResultSchemaProvider::ResultSchemaField::defaultValue() const {
   return defaultValue_;
 }
 

src/codec/test/ResultSchemaProvider.h

@@ -21,14 +21,14 @@ class ResultSchemaProvider : public meta::SchemaProviderIf {
                       bool nullable,
                       int32_t offset,
                       size_t nullFlagPos,
-                      Expression* defaultValue = nullptr,
+                      std::string defaultValue = "",
                       meta::cpp2::GeoShape = meta::cpp2::GeoShape::ANY);
 
     const char* name() const override;
     nebula::cpp2::PropertyType type() const override;
     bool nullable() const override;
     bool hasDefault() const override;
-    Expression* defaultValue() const override;
+    const std::string& defaultValue() const override;
     size_t size() const override;
     size_t offset() const override;
     size_t nullFlagPos() const override;
@@ -41,7 +41,7 @@ class ResultSchemaProvider : public meta::SchemaProviderIf {
     bool nullable_;
     int32_t offset_;
     size_t nullFlagPos_;
-    Expression* defaultValue_;
+    std::string defaultValue_;
     meta::cpp2::GeoShape geoShape_;
   };
 

src/codec/test/SchemaWriter.cpp

@@ -90,8 +90,14 @@ SchemaWriter& SchemaWriter::appendCol(folly::StringPiece name,
     nullFlagPos = numNullableFields_++;
   }
 
-  columns_.emplace_back(
-      name.toString(), type, size, nullable, offset, nullFlagPos, defaultValue, geoShape);
+  columns_.emplace_back(name.toString(),
+                        type,
+                        size,
+                        nullable,
+                        offset,
+                        nullFlagPos,
+                        defaultValue ? defaultValue->encode() : "",
+                        geoShape);
   nameIndex_.emplace(std::make_pair(hash, columns_.size() - 1));
 
   return *this;

src/common/CMakeLists.txt

@@ -28,3 +28,4 @@ nebula_add_subdirectory(ssl)
 nebula_add_subdirectory(geo)
 nebula_add_subdirectory(memory)
 nebula_add_subdirectory(id)
+nebula_add_subdirectory(log)

src/common/graph/Response.h

@@ -75,6 +75,8 @@
   X(E_CONFLICT, -2008)                                                        \
   X(E_INVALID_PARM, -2009)                                                    \
   X(E_WRONGCLUSTER, -2010)                                                    \
+  X(E_ZONE_NOT_ENOUGH, -2011)                                                 \
+  X(E_ZONE_IS_EMPTY, -2012)                                                   \
                                                                               \
   X(E_STORE_FAILURE, -2021)                                                   \
   X(E_STORE_SEGMENT_ILLEGAL, -2022)                                           \

src/common/log/CMakeLists.txt

@@ -0,0 +1,8 @@
+# Copyright (c) 2021 vesoft inc. All rights reserved.
+#
+# This source code is licensed under Apache 2.0 License.
+
+nebula_add_library(
+    log_monitor_obj OBJECT
+    LogMonitor.cpp
+)

src/common/log/LogMonitor.cpp

@@ -0,0 +1,69 @@
+/* Copyright (c) 2021 vesoft inc. All rights reserved.
+ *
+ * This source code is licensed under Apache 2.0 License.
+ */
+#include "common/log/LogMonitor.h"
+
+namespace nebula {
+
+// default min_warn is 256M, disk freebytes < 256M will change LOG_LEVEL to WARNING
+DEFINE_uint64(log_min_reserved_bytes_to_warn,
+              256 * (1UL << 20),
+              "if freebytes in logdir less than this, will change log level to WARN");
+
+// default min_error is 64M, disk freebytes < 64M will change LOG_LEVEL to ERROR
+DEFINE_uint64(log_min_reserved_bytes_to_error,
+              64 * (1UL << 20),
+              "if freebytes in logdir less than this, will change log level to ERROR");
+
+// default min_fatal is 4M, disk freebytes < 4M will change LOG_LEVEL to FATAL
+DEFINE_uint64(log_min_reserved_bytes_to_fatal,
+              4 * (1UL << 20),
+              "if freebytes in logdir less than this, will change log level to FATAL");
+
+// default check log_disk interval is 10s
+DEFINE_int32(log_disk_check_interval_secs, 10, "interval to check free space of log path");
+
+void LogMonitor::getLogDiskFreeByte() {
+  boost::system::error_code ec;
+  auto info = boost::filesystem::space(FLAGS_log_dir, ec);
+  if (!ec) {
+    freeByte_ = info.available;
+  } else {
+    LOG(WARNING) << "Get filesystem info of logdir: " << FLAGS_log_dir << " failed";
+  }
+}
+
+void LogMonitor::checkAndChangeLogLevel() {
+  getLogDiskFreeByte();
+
+  if (FLAGS_log_min_reserved_bytes_to_fatal > FLAGS_log_min_reserved_bytes_to_error ||
+      FLAGS_log_min_reserved_bytes_to_fatal > FLAGS_log_min_reserved_bytes_to_warn ||
+      FLAGS_log_min_reserved_bytes_to_error > FLAGS_log_min_reserved_bytes_to_warn) {
+    LOG(ERROR) << "Get Invalid config in LogMonitor, the LogMonitor config should be "
+               << "FLAGS_log_min_reserved_bytes_to_warn >"
+               << "FLAGS_log_min_reserved_bytes_to_error > FLAGS_log_min_reserved_bytes_to_fatal;";
+    return;
+  }
+
+  if (freeByte_ < FLAGS_log_min_reserved_bytes_to_fatal) {
+    LOG(ERROR) << "log disk freebyte is less than " << FLAGS_log_min_reserved_bytes_to_fatal
+               << ", change log level to FATAL";
+    FLAGS_minloglevel = google::GLOG_FATAL;
+  } else if (freeByte_ < FLAGS_log_min_reserved_bytes_to_error) {
+    LOG(ERROR) << "log disk freebyte is less than " << FLAGS_log_min_reserved_bytes_to_error
+               << ", change log level to ERROR";
+    FLAGS_minloglevel = google::GLOG_ERROR;
+  } else if (freeByte_ < FLAGS_log_min_reserved_bytes_to_warn) {
+    LOG(ERROR) << "log disk freebyte is less than " << FLAGS_log_min_reserved_bytes_to_warn
+               << ", change log level to WARNING";
+    FLAGS_minloglevel = google::GLOG_WARNING;
+  } else {
+    // if freeByte_ is bigger than every min_log_flag, reset the FLAGS_minloglevel to old value
+    if (FLAGS_minloglevel != oldMinLogLevel_) {
+      FLAGS_minloglevel = oldMinLogLevel_;
+    }
+  }
+}
+
+}  // namespace nebula

src/common/log/LogMonitor.h

@@ -0,0 +1,43 @@
+/* Copyright (c) 2021 vesoft inc. All rights reserved.
+ *
+ * This source code is licensed under Apache 2.0 License.
+ */
+#pragma once
+
+#include <boost/filesystem.hpp>
+#include <boost/system/error_code.hpp>
+
+#include "common/thread/GenericWorker.h"
+
+namespace nebula {
+
+DECLARE_uint64(log_min_reserved_bytes_to_warn);
+DECLARE_uint64(log_min_reserved_bytes_to_error);
+DECLARE_uint64(log_min_reserved_bytes_to_fatal);
+DECLARE_int32(log_disk_check_interval_secs);
+
+class LogMonitor {
+ public:
+  LogMonitor() : oldMinLogLevel_(FLAGS_minloglevel), freeByte_(1UL << 60) {
+    worker_ = std::make_shared<thread::GenericWorker>();
+    CHECK(worker_->start());
+    worker_->addRepeatTask(
+        FLAGS_log_disk_check_interval_secs * 1000, &LogMonitor::checkAndChangeLogLevel, this);
+  }
+
+  ~LogMonitor() {
+    worker_->stop();
+    worker_->wait();
+  }
+
+  void getLogDiskFreeByte();
+
+  void checkAndChangeLogLevel();
+
+ private:
+  int32_t oldMinLogLevel_;
+  std::shared_ptr<thread::GenericWorker> worker_;
+  std::atomic_uint64_t freeByte_;
+};
+
+}  // namespace nebula