HBASE-28600 Enable setting blockcache on-heap sizes in bytes (#6422) (#6529)

Signed-off-by: Nick Dimiduk <[email protected]>
Co-authored-by: JinHyuk Kim <[email protected]>

Author: ndimiduk

hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java

@@ -1012,6 +1012,11 @@ public enum OperationStatusCode {
 
   public static final float HFILE_BLOCK_CACHE_SIZE_DEFAULT = 0.4f;
 
+  /**
+   * Configuration key for the memory size of the block cache
+   */
+  public static final String HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY = "hfile.block.cache.memory.size";
+
   /**
    * Configuration key for setting the fix size of the block size, default do nothing and it should
    * be explicitly set by user or only used within ClientSideRegionScanner. if it's set less than

hbase-common/src/main/resources/hbase-default.xml

@@ -1010,6 +1010,14 @@ possible configurations would overwhelm and obscure the important.
         Set to 0 to disable but it's not recommended; you need at least
         enough cache to hold the storefile indices.</description>
   </property>
+  <property>
+    <name>hfile.block.cache.memory.size</name>
+    <value></value>
+    <description>Defines the maximum heap memory allocated for the HFile block cache,
+      specified in bytes or human-readable formats like '10m' for megabytes or '10g' for gigabytes.
+      This configuration allows setting an absolute memory size instead of a percentage of the maximum heap.
+      Takes precedence over hfile.block.cache.size if both are specified.</description>
+  </property>
   <property>
       <name>hfile.block.index.cacheonwrite</name>
       <value>false</value>

hbase-server/src/main/java/org/apache/hadoop/hbase/io/util/MemorySizeUtil.java

@@ -21,6 +21,7 @@
 import java.lang.management.MemoryType;
 import java.lang.management.MemoryUsage;
 import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.conf.StorageUnit;
 import org.apache.hadoop.hbase.HConstants;
 import org.apache.hadoop.hbase.regionserver.MemStoreLAB;
 import org.apache.hadoop.hbase.util.Pair;
@@ -93,11 +94,16 @@ public static void checkForClusterFreeHeapMemoryLimit(Configuration conf) {
     ) {
       throw new RuntimeException("Current heap configuration for MemStore and BlockCache exceeds "
         + "the threshold required for successful cluster operation. "
-        + "The combined value cannot exceed 0.8. Please check "
-        + "the settings for hbase.regionserver.global.memstore.size and "
-        + "hfile.block.cache.size in your configuration. "
-        + "hbase.regionserver.global.memstore.size is " + globalMemstoreSize
-        + " hfile.block.cache.size is " + blockCacheUpperLimit);
+        + "The combined value cannot exceed 0.8. Please check " + "the settings for "
+        + MEMSTORE_SIZE_KEY + " and either " + HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY + " or "
+        + HConstants.HFILE_BLOCK_CACHE_SIZE_KEY + " in your configuration. " + MEMSTORE_SIZE_KEY
+        + "=" + globalMemstoreSize + ", " + HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY + "="
+        + conf.get(HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY) + ", "
+        + HConstants.HFILE_BLOCK_CACHE_SIZE_KEY + "="
+        + conf.get(HConstants.HFILE_BLOCK_CACHE_SIZE_KEY) + ". (Note: If both "
+        + HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY + " and "
+        + HConstants.HFILE_BLOCK_CACHE_SIZE_KEY + " are set, " + "the system will use "
+        + HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY + ")");
     }
   }
 
@@ -195,10 +201,30 @@ public static long getOnheapGlobalMemStoreSize(Configuration conf) {
    * Retrieve configured size for on heap block cache as percentage of total heap.
    */
   public static float getBlockCacheHeapPercent(final Configuration conf) {
-    // L1 block cache is always on heap
-    float l1CachePercent = conf.getFloat(HConstants.HFILE_BLOCK_CACHE_SIZE_KEY,
+    // Check if an explicit block cache size is configured.
+    long l1CacheSizeInBytes = getBlockCacheSizeInBytes(conf);
+    if (l1CacheSizeInBytes > 0) {
+      final MemoryUsage usage = safeGetHeapMemoryUsage();
+      return usage == null ? 0 : (float) l1CacheSizeInBytes / usage.getMax();
+    }
+
+    return conf.getFloat(HConstants.HFILE_BLOCK_CACHE_SIZE_KEY,
       HConstants.HFILE_BLOCK_CACHE_SIZE_DEFAULT);
-    return l1CachePercent;
+  }
+
+  /**
+   * Retrieve an explicit block cache size in bytes in the configuration.
+   * @param conf used to read cache configs
+   * @return the number of bytes to use for LRU, negative if disabled.
+   * @throws IllegalArgumentException if HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY format is invalid
+   */
+  public static long getBlockCacheSizeInBytes(Configuration conf) {
+    final String key = HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY;
+    try {
+      return Long.parseLong(conf.get(key));
+    } catch (NumberFormatException e) {
+      return (long) conf.getStorageSize(key, -1, StorageUnit.BYTES);
+    }
   }
 
   /**
@@ -207,27 +233,30 @@ public static float getBlockCacheHeapPercent(final Configuration conf) {
    * @throws IllegalArgumentException if HFILE_BLOCK_CACHE_SIZE_KEY is > 1.0
    */
   public static long getOnHeapCacheSize(final Configuration conf) {
-    float cachePercentage = conf.getFloat(HConstants.HFILE_BLOCK_CACHE_SIZE_KEY,
-      HConstants.HFILE_BLOCK_CACHE_SIZE_DEFAULT);
+    final float cachePercentage = getBlockCacheHeapPercent(conf);
     if (cachePercentage <= 0.0001f) {
       return -1;
     }
     if (cachePercentage > 1.0) {
       throw new IllegalArgumentException(
         HConstants.HFILE_BLOCK_CACHE_SIZE_KEY + " must be between 0.0 and 1.0, and not > 1.0");
     }
-    long max = -1L;
+
     final MemoryUsage usage = safeGetHeapMemoryUsage();
-    if (usage != null) {
-      max = usage.getMax();
+    if (usage == null) {
+      return -1;
     }
+    final long heapMax = usage.getMax();
     float onHeapCacheFixedSize =
       (float) conf.getLong(HConstants.HFILE_ONHEAP_BLOCK_CACHE_FIXED_SIZE_KEY,
-        HConstants.HFILE_ONHEAP_BLOCK_CACHE_FIXED_SIZE_DEFAULT) / max;
+        HConstants.HFILE_ONHEAP_BLOCK_CACHE_FIXED_SIZE_DEFAULT) / heapMax;
     // Calculate the amount of heap to give the heap.
-    return (onHeapCacheFixedSize > 0 && onHeapCacheFixedSize < cachePercentage)
-      ? (long) (max * onHeapCacheFixedSize)
-      : (long) (max * cachePercentage);
+    if (onHeapCacheFixedSize > 0 && onHeapCacheFixedSize < cachePercentage) {
+      return (long) (heapMax * onHeapCacheFixedSize);
+    } else {
+      final long cacheSizeInBytes = getBlockCacheSizeInBytes(conf);
+      return cacheSizeInBytes > 0 ? cacheSizeInBytes : (long) (heapMax * cachePercentage);
+    }
   }
 
   /**
@@ -243,5 +272,4 @@ public static long getBucketCacheSize(final Configuration conf) {
     }
     return (long) (bucketCacheSize * 1024 * 1024);
   }
-
 }

hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/HeapMemoryManager.java

@@ -17,6 +17,7 @@
  */
 package org.apache.hadoop.hbase.regionserver;
 
+import static org.apache.hadoop.hbase.HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY;
 import static org.apache.hadoop.hbase.HConstants.HFILE_BLOCK_CACHE_SIZE_KEY;
 
 import java.lang.management.MemoryUsage;
@@ -128,8 +129,7 @@ private ResizableBlockCache toResizableBlockCache(BlockCache blockCache) {
   private boolean doInit(Configuration conf) {
     boolean tuningEnabled = true;
     globalMemStorePercent = MemorySizeUtil.getGlobalMemStoreHeapPercent(conf, false);
-    blockCachePercent =
-      conf.getFloat(HFILE_BLOCK_CACHE_SIZE_KEY, HConstants.HFILE_BLOCK_CACHE_SIZE_DEFAULT);
+    blockCachePercent = MemorySizeUtil.getBlockCacheHeapPercent(conf);
     MemorySizeUtil.checkForClusterFreeHeapMemoryLimit(conf);
     // Initialize max and min range for memstore heap space
     globalMemStorePercentMinRange =
@@ -160,16 +160,20 @@ private boolean doInit(Configuration conf) {
     blockCachePercentMinRange = conf.getFloat(BLOCK_CACHE_SIZE_MIN_RANGE_KEY, blockCachePercent);
     blockCachePercentMaxRange = conf.getFloat(BLOCK_CACHE_SIZE_MAX_RANGE_KEY, blockCachePercent);
     if (blockCachePercent < blockCachePercentMinRange) {
-      LOG.warn("Setting " + BLOCK_CACHE_SIZE_MIN_RANGE_KEY + " to " + blockCachePercent
-        + ", same value as " + HFILE_BLOCK_CACHE_SIZE_KEY
-        + " because supplied value greater than initial block cache size.");
+      LOG.warn(
+        "Setting {} to {} (lookup order: {} -> {}), "
+          + "because supplied value greater than initial block cache size.",
+        BLOCK_CACHE_SIZE_MIN_RANGE_KEY, blockCachePercent, HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY,
+        HFILE_BLOCK_CACHE_SIZE_KEY);
       blockCachePercentMinRange = blockCachePercent;
       conf.setFloat(BLOCK_CACHE_SIZE_MIN_RANGE_KEY, blockCachePercentMinRange);
     }
     if (blockCachePercent > blockCachePercentMaxRange) {
-      LOG.warn("Setting " + BLOCK_CACHE_SIZE_MAX_RANGE_KEY + " to " + blockCachePercent
-        + ", same value as " + HFILE_BLOCK_CACHE_SIZE_KEY
-        + " because supplied value less than initial block cache size.");
+      LOG.warn(
+        "Setting {} to {} (lookup order: {} -> {}), "
+          + "because supplied value less than initial block cache size.",
+        BLOCK_CACHE_SIZE_MAX_RANGE_KEY, blockCachePercent, HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY,
+        HFILE_BLOCK_CACHE_SIZE_KEY);
       blockCachePercentMaxRange = blockCachePercent;
       conf.setFloat(BLOCK_CACHE_SIZE_MAX_RANGE_KEY, blockCachePercentMaxRange);
     }

hbase-server/src/test/java/org/apache/hadoop/hbase/io/hfile/TestCacheConfig.java

@@ -396,6 +396,14 @@ public void testGetOnHeapCacheSize() {
     long onHeapCacheSize = MemorySizeUtil.getOnHeapCacheSize(copyConf);
     assertEquals(null, copyConf.get(HConstants.HFILE_ONHEAP_BLOCK_CACHE_FIXED_SIZE_KEY));
     assertTrue(onHeapCacheSize > 0 && onHeapCacheSize != fixedSize);
+    // when HBASE_BLOCK_CACHE_MEMORY_SIZE is set in number
+    copyConf.setLong(HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY, 3 * 1024 * 1024);
+    onHeapCacheSize = MemorySizeUtil.getOnHeapCacheSize(copyConf);
+    assertEquals(3 * 1024 * 1024, onHeapCacheSize);
+    // when HBASE_BLOCK_CACHE_MEMORY_SIZE is set in human-readable format
+    copyConf.set(HConstants.HFILE_BLOCK_CACHE_MEMORY_SIZE_KEY, "2m");
+    onHeapCacheSize = MemorySizeUtil.getOnHeapCacheSize(copyConf);
+    assertEquals(2 * 1024 * 1024, onHeapCacheSize);
     // when HBASE_BLOCK_CACHE_FIXED_SIZE_KEY is set, it will be a fixed size
     copyConf.setLong(HConstants.HFILE_ONHEAP_BLOCK_CACHE_FIXED_SIZE_KEY, fixedSize);
     onHeapCacheSize = MemorySizeUtil.getOnHeapCacheSize(copyConf);