Add interface for rate of found elements in the dictionaries

- IDictionary abstraction
- skeleton implementation into each dictionary
- system.dictionaries.found_rate
- documentation changes

docs/en/operations/system-tables/dictionaries.md

@@ -21,6 +21,7 @@ Columns:
 -   `bytes_allocated` ([UInt64](../../sql-reference/data-types/int-uint.md#uint-ranges)) — Amount of RAM allocated for the dictionary.
 -   `query_count` ([UInt64](../../sql-reference/data-types/int-uint.md#uint-ranges)) — Number of queries since the dictionary was loaded or since the last successful reboot.
 -   `hit_rate` ([Float64](../../sql-reference/data-types/float.md)) — For cache dictionaries, the percentage of uses for which the value was in the cache.
+-   `found_rate` ([Float64](../../sql-reference/data-types/float.md)) — The percentage of uses for which the value was found.
 -   `element_count` ([UInt64](../../sql-reference/data-types/int-uint.md#uint-ranges)) — Number of items stored in the dictionary.
 -   `load_factor` ([Float64](../../sql-reference/data-types/float.md)) — Percentage filled in the dictionary (for a hashed dictionary, the percentage filled in the hash table).
 -   `source` ([String](../../sql-reference/data-types/string.md)) — Text describing the [data source](../../sql-reference/dictionaries/external-dictionaries/external-dicts-dict-sources.md) for the dictionary.
@@ -60,4 +61,4 @@ SELECT * FROM system.dictionaries
 └──────────┴──────┴────────┴─────────────┴──────┴────────┴──────────────────────────────────────┴─────────────────────┴─────────────────┴─────────────┴──────────┴───────────────┴───────────────────────┴────────────────────────────┴──────────────┴──────────────┴─────────────────────┴──────────────────────────────┘───────────────────────┴────────────────┘
 ```
 
-[Original article](https://clickhouse.tech/docs/en/operations/system_tables/dictionaries) <!--hide-->
+[Original article](https://clickhouse.tech/docs/en/operations/system_tables/dictionaries) <!--hide-->

src/Dictionaries/CacheDictionary.h

@@ -75,6 +75,14 @@ public:
 
     size_t getQueryCount() const override { return query_count.load(std::memory_order_relaxed); }
 
+    double getFoundRate() const override
+    {
+        size_t queries = query_count.load(std::memory_order_relaxed);
+        if (!queries)
+            return 0;
+        return static_cast<double>(found_count.load(std::memory_order_relaxed)) / queries;
+    }
+
     double getHitRate() const override
     {
         return static_cast<double>(hit_count.load(std::memory_order_acquire)) / query_count.load(std::memory_order_relaxed);
@@ -204,6 +212,7 @@ private:
 
     mutable std::atomic<size_t> hit_count{0};
     mutable std::atomic<size_t> query_count{0};
+    mutable std::atomic<size_t> found_count{0};
 
 };
 

src/Dictionaries/DirectDictionary.h

@@ -42,6 +42,14 @@ public:
 
     size_t getQueryCount() const override { return query_count.load(std::memory_order_relaxed); }
 
+    double getFoundRate() const override
+    {
+        size_t queries = query_count.load(std::memory_order_relaxed);
+        if (!queries)
+            return 0;
+        return static_cast<double>(found_count.load(std::memory_order_relaxed)) / queries;
+    }
+
     double getHitRate() const override { return 1.0; }
 
     size_t getElementCount() const override { return 0; }
@@ -101,6 +109,7 @@ private:
     const DictionaryLifetime dict_lifetime;
 
     mutable std::atomic<size_t> query_count{0};
+    mutable std::atomic<size_t> found_count{0};
 };
 
 extern template class DirectDictionary<DictionaryKeyType::simple>;

src/Dictionaries/FlatDictionary.h

@@ -47,6 +47,14 @@ public:
 
     size_t getQueryCount() const override { return query_count.load(std::memory_order_relaxed); }
 
+    double getFoundRate() const override
+    {
+        size_t queries = query_count.load(std::memory_order_relaxed);
+        if (!queries)
+            return 0;
+        return static_cast<double>(found_count.load(std::memory_order_relaxed)) / queries;
+    }
+
     double getHitRate() const override { return 1.0; }
 
     size_t getElementCount() const override { return element_count; }
@@ -183,6 +191,7 @@ private:
     size_t element_count = 0;
     size_t bucket_count = 0;
     mutable std::atomic<size_t> query_count{0};
+    mutable std::atomic<size_t> found_count{0};
 
     BlockPtr update_field_loaded_block;
 };

src/Dictionaries/HashedDictionary.h

@@ -59,6 +59,14 @@ public:
 
     size_t getQueryCount() const override { return query_count.load(std::memory_order_relaxed); }
 
+    double getFoundRate() const override
+    {
+        size_t queries = query_count.load(std::memory_order_relaxed);
+        if (!queries)
+            return 0;
+        return static_cast<double>(found_count.load(std::memory_order_relaxed)) / queries;
+    }
+
     double getHitRate() const override { return 1.0; }
 
     size_t getElementCount() const override { return element_count; }
@@ -219,6 +227,7 @@ private:
     size_t element_count = 0;
     size_t bucket_count = 0;
     mutable std::atomic<size_t> query_count{0};
+    mutable std::atomic<size_t> found_count{0};
 
     BlockPtr update_field_loaded_block;
     Arena complex_key_arena;

src/Dictionaries/IDictionary.h

@@ -90,6 +90,8 @@ struct IDictionary : public IExternalLoadable
 
     virtual size_t getQueryCount() const = 0;
 
+    virtual double getFoundRate() const = 0;
+
     virtual double getHitRate() const = 0;
 
     virtual size_t getElementCount() const = 0;

src/Dictionaries/IPAddressDictionary.h

@@ -38,6 +38,14 @@ public:
 
     size_t getQueryCount() const override { return query_count.load(std::memory_order_relaxed); }
 
+    double getFoundRate() const override
+    {
+        size_t queries = query_count.load(std::memory_order_relaxed);
+        if (!queries)
+            return 0;
+        return static_cast<double>(found_count.load(std::memory_order_relaxed)) / queries;
+    }
+
     double getHitRate() const override { return 1.0; }
 
     size_t getElementCount() const override { return element_count; }
@@ -202,6 +210,7 @@ private:
     size_t element_count = 0;
     size_t bucket_count = 0;
     mutable std::atomic<size_t> query_count{0};
+    mutable std::atomic<size_t> found_count{0};
 
     Poco::Logger * logger;
 };

src/Dictionaries/PolygonDictionary.h

@@ -63,6 +63,14 @@ public:
 
     size_t getQueryCount() const override { return query_count.load(std::memory_order_relaxed); }
 
+    double getFoundRate() const override
+    {
+        size_t queries = query_count.load(std::memory_order_relaxed);
+        if (!queries)
+            return 0;
+        return static_cast<double>(found_count.load(std::memory_order_relaxed)) / queries;
+    }
+
     double getHitRate() const override { return 1.0; }
 
     size_t getElementCount() const override { return attributes.empty() ? 0 : attributes.front()->size(); }
@@ -141,6 +149,7 @@ private:
 
     size_t bytes_allocated = 0;
     mutable std::atomic<size_t> query_count{0};
+    mutable std::atomic<size_t> found_count{0};
 
     /** Since the original data may have been in the form of multi-polygons, an id is stored for each single polygon
      *  corresponding to the row in which any other attributes for this entry are located.

src/Dictionaries/RangeHashedDictionary.h

@@ -32,6 +32,14 @@ public:
 
     size_t getQueryCount() const override { return query_count.load(std::memory_order_relaxed); }
 
+    double getFoundRate() const override
+    {
+        size_t queries = query_count.load(std::memory_order_relaxed);
+        if (!queries)
+            return 0;
+        return static_cast<double>(found_count.load(std::memory_order_relaxed)) / queries;
+    }
+
     double getHitRate() const override { return 1.0; }
 
     size_t getElementCount() const override { return element_count; }
@@ -201,6 +209,7 @@ private:
     size_t element_count = 0;
     size_t bucket_count = 0;
     mutable std::atomic<size_t> query_count{0};
+    mutable std::atomic<size_t> found_count{0};
 };
 
 }

src/Storages/System/StorageSystemDictionaries.cpp

@@ -37,6 +37,7 @@ NamesAndTypesList StorageSystemDictionaries::getNamesAndTypes()
         {"bytes_allocated", std::make_shared<DataTypeUInt64>()},
         {"query_count", std::make_shared<DataTypeUInt64>()},
         {"hit_rate", std::make_shared<DataTypeFloat64>()},
+        {"found_rate", std::make_shared<DataTypeFloat64>()},
         {"element_count", std::make_shared<DataTypeUInt64>()},
         {"load_factor", std::make_shared<DataTypeFloat64>()},
         {"source", std::make_shared<DataTypeString>()},
@@ -113,6 +114,7 @@ void StorageSystemDictionaries::fillData(MutableColumns & res_columns, ContextPt
             res_columns[i++]->insert(dict_ptr->getBytesAllocated());
             res_columns[i++]->insert(dict_ptr->getQueryCount());
             res_columns[i++]->insert(dict_ptr->getHitRate());
+            res_columns[i++]->insert(dict_ptr->getFoundRate());
             res_columns[i++]->insert(dict_ptr->getElementCount());
             res_columns[i++]->insert(dict_ptr->getLoadFactor());
             res_columns[i++]->insert(dict_ptr->getSource()->toString());
@@ -125,7 +127,7 @@ void StorageSystemDictionaries::fillData(MutableColumns & res_columns, ContextPt
         }
         else
         {
-            for (size_t j = 0; j != 8; ++j) // Number of empty fields if dict_ptr is null
+            for (size_t j = 0; j != 9; ++j) // Number of empty fields if dict_ptr is null
                 res_columns[i++]->insertDefault();
         }