How would you implement an LRU cache in Java?

Question

Please don't say EHCache or OSCache, etc. Assume for purposes of this question that I want to implement my own using just the SDK (learning by doing). Given that the cache will be used in a multithreaded environment, which datastructures would you use? I've already implemented one using LinkedHashMap and Collections#synchronizedMap, but I'm curious if any of the new concurrent collections would be better candidates.

UPDATE: I was just reading through Yegge's latest when I found this nugget:

If you need constant-time access and want to maintain the insertion order, you can't do better than a LinkedHashMap, a truly wonderful data structure. The only way it could possibly be more wonderful is if there were a concurrent version. But alas.

I was thinking almost exactly the same thing before I went with the LinkedHashMap + Collections#synchronizedMap implementation I mentioned above. Nice to know I hadn't just overlooked something.

Based on the answers so far, it sounds like my best bet for a highly concurrent LRU would be to extend ConcurrentHashMap using some of the same logic that LinkedHashMap uses.

Answer 1

I like lots of these suggestions, but for now I think I'll stick with LinkedHashMap + Collections.synchronizedMap. If I do revisit this in the future, I'll probably work on extending ConcurrentHashMap in the same way LinkedHashMap extends HashMap.

UPDATE:

By request, here's the gist of my current implementation.

private class LruCache<A, B> extends LinkedHashMap<A, B> {
    private final int maxEntries;

    public LruCache(final int maxEntries) {
        super(maxEntries + 1, 1.0f, true);
        this.maxEntries = maxEntries;
    }

    /**
     * Returns <tt>true</tt> if this <code>LruCache</code> has more entries than the maximum specified when it was
     * created.
     *
     * <p>
     * This method <em>does not</em> modify the underlying <code>Map</code>; it relies on the implementation of
     * <code>LinkedHashMap</code> to do that, but that behavior is documented in the JavaDoc for
     * <code>LinkedHashMap</code>.
     * </p>
     *
     * @param eldest
     *            the <code>Entry</code> in question; this implementation doesn't care what it is, since the
     *            implementation is only dependent on the size of the cache
     * @return <tt>true</tt> if the oldest
     * @see java.util.LinkedHashMap#removeEldestEntry(Map.Entry)
     */
    @Override
    protected boolean removeEldestEntry(final Map.Entry<A, B> eldest) {
        return super.size() > maxEntries;
    }
}

Map<String, String> example = Collections.synchronizedMap(new LruCache<String, String>(CACHE_SIZE));

Answer 2

Hope this helps .

import java.util.*;
public class Lru {

public static <K,V> Map<K,V> lruCache(final int maxSize) {
    return new LinkedHashMap<K, V>(maxSize*4/3, 0.75f, true) {

        private static final long serialVersionUID = -3588047435434569014L;

        @Override
        protected boolean removeEldestEntry(Map.Entry<K, V> eldest) {
            return size() > maxSize;
        }
    };
 }
 public static void main(String[] args ) {
    Map<Object, Object> lru = Lru.lruCache(2);      
    lru.put("1", "1");
    lru.put("2", "2");
    lru.put("3", "3");
    System.out.println(lru);
}
}

Answer 3

Here is my short implementation, please criticize or improve it!

package util.collection;

import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ConcurrentLinkedQueue;

/**
 * Limited size concurrent cache map implementation.<br/>
 * LRU: Least Recently Used.<br/>
 * If you add a new key-value pair to this cache after the maximum size has been exceeded,
 * the oldest key-value pair will be removed before adding.
 */

public class ConcurrentLRUCache<Key, Value> {

private final int maxSize;
private int currentSize = 0;

private ConcurrentHashMap<Key, Value> map;
private ConcurrentLinkedQueue<Key> queue;

public ConcurrentLRUCache(final int maxSize) {
    this.maxSize = maxSize;
    map = new ConcurrentHashMap<Key, Value>(maxSize);
    queue = new ConcurrentLinkedQueue<Key>();
}

private synchronized void freeSpace() {
    Key key = queue.poll();
    if (null != key) {
        map.remove(key);
        currentSize = map.size();
    }
}

public void put(Key key, Value val) {
    if (map.containsKey(key)) {// just heat up that item
        put(key, val);
        return;
    }
    while (currentSize >= maxSize) {
        freeSpace();
    }
    synchronized(this) {
        queue.add(key);
        map.put(key, val);
        currentSize++;
    }
}

public Value get(Key key) {
    return map.get(key);
}
}

Answer 4

Wanted to add comment to the answer given by Hank but some how I am not able to - please treat it as comment

LinkedHashMap maintains access order as well based on parameter passed in its constructor It keeps doubly lined list to maintain order (See LinkedHashMap.Entry)

@Pacerier it is correct that LinkedHashMap keeps same order while iteration if element is added again but that is only in case of insertion order mode.

this is what I found in java docs of LinkedHashMap.Entry object

    /**
     * This method is invoked by the superclass whenever the value
     * of a pre-existing entry is read by Map.get or modified by Map.set.
     * If the enclosing Map is access-ordered, it moves the entry
     * to the end of the list; otherwise, it does nothing.
     */
    void recordAccess(HashMap<K,V> m) {
        LinkedHashMap<K,V> lm = (LinkedHashMap<K,V>)m;
        if (lm.accessOrder) {
            lm.modCount++;
            remove();
            addBefore(lm.header);
        }
    }

this method takes care of moving recently accessed element to end of the list. So all in all LinkedHashMap is best data structure for implementing LRUCache.

Answer 5

Another thought and even a simple implementation using LinkedHashMap collection of Java.

LinkedHashMap provided method removeEldestEntry and which can be overridden in the way mentioned in example. By default implementation of this collection structure is false. If its true and size of this structure goes beyond the initial capacity than eldest or older elements will be removed.

We can have a pageno and page content in my case pageno is integer and pagecontent i have kept page number values string.

import java.util.LinkedHashMap;
import java.util.Map;

/**
 * @author Deepak Singhvi
 *
 */
public class LRUCacheUsingLinkedHashMap {


     private static int CACHE_SIZE = 3;
     public static void main(String[] args) {
        System.out.println(" Pages for consideration : 2, 1, 0, 2, 8, 2, 4,99");
        System.out.println("----------------------------------------------\n");


// accessOrder is true, so whenever any page gets changed or accessed,    // its order will change in the map, 
              LinkedHashMap<Integer,String> lruCache = new              
                 LinkedHashMap<Integer,String>(CACHE_SIZE, .75F, true) {

           private static final long serialVersionUID = 1L;

           protected boolean removeEldestEntry(Map.Entry<Integer,String>                           

                     eldest) {
                          return size() > CACHE_SIZE;
                     }

                };

  lruCache.put(2, "2");
  lruCache.put(1, "1");
  lruCache.put(0, "0");
  System.out.println(lruCache + "  , After first 3 pages in cache");
  lruCache.put(2, "2");
  System.out.println(lruCache + "  , Page 2 became the latest page in the cache");
  lruCache.put(8, "8");
  System.out.println(lruCache + "  , Adding page 8, which removes eldest element 2 ");
  lruCache.put(2, "2");
  System.out.println(lruCache+ "  , Page 2 became the latest page in the cache");
  lruCache.put(4, "4");
  System.out.println(lruCache+ "  , Adding page 4, which removes eldest element 1 ");
  lruCache.put(99, "99");
  System.out.println(lruCache + " , Adding page 99, which removes eldest element 8 ");

     }

}

Result of above code execution is as follows:

 Pages for consideration : 2, 1, 0, 2, 8, 2, 4,99
--------------------------------------------------
    {2=2, 1=1, 0=0}  , After first 3 pages in cache
    {2=2, 1=1, 0=0}  , Page 2 became the latest page in the cache
    {1=1, 0=0, 8=8}  , Adding page 8, which removes eldest element 2 
    {0=0, 8=8, 2=2}  , Page 2 became the latest page in the cache
    {8=8, 2=2, 4=4}  , Adding page 4, which removes eldest element 1 
    {2=2, 4=4, 99=99} , Adding page 99, which removes eldest element 8 

Answer 6

Here is my tested best performing concurrent LRU cache implementation without any synchronized block:

public class ConcurrentLRUCache<Key, Value> {

private final int maxSize;

private ConcurrentHashMap<Key, Value> map;
private ConcurrentLinkedQueue<Key> queue;

public ConcurrentLRUCache(final int maxSize) {
    this.maxSize = maxSize;
    map = new ConcurrentHashMap<Key, Value>(maxSize);
    queue = new ConcurrentLinkedQueue<Key>();
}

/**
 * @param key - may not be null!
 * @param value - may not be null!
 */
public void put(final Key key, final Value value) {
    if (map.containsKey(key)) {
        queue.remove(key); // remove the key from the FIFO queue
    }

    while (queue.size() >= maxSize) {
        Key oldestKey = queue.poll();
        if (null != oldestKey) {
            map.remove(oldestKey);
        }
    }
    queue.add(key);
    map.put(key, value);
}

/**
 * @param key - may not be null!
 * @return the value associated to the given key or null
 */
public Value get(final Key key) {
    return map.get(key);
}

}