Extend nodetool tablestats for dictionary memory usage

patch by Stefan Miklosovic; reviewed by Yifan Cai for CASSANDRA-20940

Author: smiklosovic

CHANGES.txt

@@ -1,4 +1,5 @@
 5.1
+ * Extend nodetool tablestats for dictionary memory usage (CASSANDRA-20940)
  * Introduce separate GCInspector thresholds for concurrent GC events (CASSANDRA-20980)
  * Reduce contention in MemtableAllocator.allocate (CASSANDRA-20226)
  * Add export, list, import sub-commands for nodetool compressiondictionary (CASSANDRA-20941)

src/java/org/apache/cassandra/db/compression/CompressionDictionary.java

@@ -56,6 +56,15 @@ public interface CompressionDictionary extends AutoCloseable
      */
     int checksum();
 
+    /**
+     * Get memory occupied of this dictionary as a whole.
+     * Use for metrics exposing used memory in total. The value
+     * return by this method is a best-effort estimation.
+     *
+     * @return memory occuppied by this compression dictionary, in bytes
+     */
+    int estimatedOccupiedMemoryBytes();
+
     /**
      * Get the kind of the compression algorithm
      *

src/java/org/apache/cassandra/db/compression/CompressionDictionaryCache.java

@@ -19,6 +19,7 @@
 package org.apache.cassandra.db.compression;
 
 import java.time.Duration;
+import java.util.concurrent.atomic.AtomicInteger;
 import java.util.concurrent.atomic.AtomicReference;
 import javax.annotation.Nullable;
 
@@ -112,6 +113,14 @@ public void add(@Nullable CompressionDictionary compressionDictionary)
         }
     }
 
+    @Override
+    public long cachedDictionariesMemoryUsed()
+    {
+        AtomicInteger value = new AtomicInteger();
+        cache.asMap().forEach((key, dict) -> value.addAndGet(dict.estimatedOccupiedMemoryBytes()));
+        return value.get();
+    }
+
     @Override
     public synchronized void close()
     {

src/java/org/apache/cassandra/db/compression/CompressionDictionaryManager.java

@@ -192,6 +192,12 @@ public void add(@Nullable CompressionDictionary compressionDictionary)
         cache.add(compressionDictionary);
     }
 
+    @Override
+    public long cachedDictionariesMemoryUsed()
+    {
+        return cache.cachedDictionariesMemoryUsed();
+    }
+
     @Override
     public void onNewDictionaryTrained(CompressionDictionary.DictId dictionaryId)
     {

src/java/org/apache/cassandra/db/compression/ICompressionDictionaryCache.java

@@ -53,4 +53,11 @@ public interface ICompressionDictionaryCache extends AutoCloseable
      * @param compressionDictionary the compression dictionary to cache, may be null
      */
     void add(@Nullable CompressionDictionary compressionDictionary);
+
+    /**
+     * Gives number of bytes cached compression dictionaries occupy in this cache.
+     *
+     * @return number of bytes cached dictionaries occupy
+     */
+    long cachedDictionariesMemoryUsed();
 }

src/java/org/apache/cassandra/db/compression/ZstdCompressionDictionary.java

@@ -86,6 +86,16 @@ public int checksum()
         return checksum;
     }
 
+    @Override
+    public int estimatedOccupiedMemoryBytes()
+    {
+        int occupied = rawDictionary.length;
+        occupied += dictDecompress != null ? rawDictionary.length : 0;
+        occupied += zstdDictCompressPerLevel.size() * rawDictionary.length;
+
+        return occupied;
+    }
+
     @Override
     public boolean equals(Object o)
     {

src/java/org/apache/cassandra/metrics/TableMetrics.java

@@ -45,6 +45,7 @@
 import org.apache.cassandra.config.DatabaseDescriptor;
 import org.apache.cassandra.db.ColumnFamilyStore;
 import org.apache.cassandra.db.Keyspace;
+import org.apache.cassandra.db.compression.CompressionDictionaryManager;
 import org.apache.cassandra.db.lifecycle.SSTableSet;
 import org.apache.cassandra.db.lifecycle.View;
 import org.apache.cassandra.db.memtable.Memtable;
@@ -154,6 +155,8 @@ public class TableMetrics
     public final Gauge<Long> maxPartitionSize;
     /** Size of the smallest compacted partition */
     public final Gauge<Long> meanPartitionSize;
+    /** Memory usage of cached compression dictionaries */
+    public final Gauge<Long> compressionDictionariesMemoryUsed;
     /** Off heap memory used by compression meta data*/
     public final Gauge<Long> compressionMetadataOffHeapMemoryUsed;
     /** Tombstones scanned in queries on this CF */
@@ -771,6 +774,19 @@ public Long getValue()
                 return count > 0 ? sum / count : 0;
             }
         });
+        compressionDictionariesMemoryUsed = createTableGauge("CompressionDictionariesMemoryUsed", new Gauge<Long>()
+        {
+            @Override
+            public Long getValue()
+            {
+                CompressionDictionaryManager compressionDictionaryManager = cfs.compressionDictionaryManager();
+                if (compressionDictionaryManager != null)
+                    return compressionDictionaryManager.cachedDictionariesMemoryUsed();
+                else
+                    return 0L;
+            }
+        });
+
         compressionMetadataOffHeapMemoryUsed = createTableGauge("CompressionMetadataOffHeapMemoryUsed", new Gauge<Long>()
         {
             public Long getValue()

src/java/org/apache/cassandra/tools/NodeProbe.java

@@ -2088,6 +2088,7 @@ else if (!Strings.isNullOrEmpty(ks))
                 case "BloomFilterFalseRatio":
                 case "BloomFilterOffHeapMemoryUsed":
                 case "IndexSummaryOffHeapMemoryUsed":
+                case "CompressionDictionariesMemoryUsed":
                 case "CompressionMetadataOffHeapMemoryUsed":
                 case "CompressionRatio":
                 case "EstimatedColumnCountHistogram":

src/java/org/apache/cassandra/tools/nodetool/TableStats.java

@@ -58,7 +58,7 @@ public class TableStats extends AbstractCommand
                         + "average_tombstones_per_slice_last_five_minutes, bloom_filter_false_positives, "
                         + "bloom_filter_false_ratio, bloom_filter_off_heap_memory_used, bloom_filter_space_used, "
                         + "compacted_partition_maximum_bytes, compacted_partition_mean_bytes, "
-                        + "compacted_partition_minimum_bytes, compression_metadata_off_heap_memory_used, "
+                        + "compacted_partition_minimum_bytes, compression_dictionaries_memory_used, compression_metadata_off_heap_memory_used, "
                         + "full_name, index_summary_off_heap_memory_used, local_read_count, "
                         + "local_read_latency_ms, local_write_latency_ms, "
                         + "maximum_live_cells_per_slice_last_five_minutes, "

src/java/org/apache/cassandra/tools/nodetool/stats/StatsTable.java

@@ -57,6 +57,8 @@ public class StatsTable
     public String bloomFilterOffHeapMemoryUsed;
     public boolean indexSummaryOffHeapUsed = false;
     public String indexSummaryOffHeapMemoryUsed;
+    public boolean compressionDictionariesUsed = false;
+    public String compressionDictionariesMemoryUsed;
     public boolean compressionMetadataOffHeapUsed = false;
     public String compressionMetadataOffHeapMemoryUsed;
     public long compactedPartitionMinimumBytes;