I have three different threads which creates three different objects to read/manipulate some data which is common for all the threads. Now, I need to ensure that we are giving an access only to one thread at a time.
The example goes something like this.
public interface CommonData {
public void addData(); // adds data to the cache
public void getDataAccessKey(); // Key that will be common across different threads for each data type
}
/*
* Singleton class
*/
public class CommonDataCache() {
private final Map dataMap = new HashMap(); // this takes keys and values as custom objects
}
The implementation class of the interface would look like this
class CommonDataImpl implements CommonData {
private String key;
public CommonDataImpl1(String key) {
this.key = key;
}
public void addData() {
// access the singleton cache class and add
}
public void getDataAccessKey() {
return key;
}
}
Each thread will be invoked as follows:
CommonData data = new CommonDataImpl("Key1");
new Thread(() -> data.addData()).start();
CommonData data1 = new CommonDataImpl("Key1");
new Thread(() -> data1.addData()).start();
CommonData data2 = new CommonDataImpl("Key1");
new Thread(() -> data2.addData()).start();
Now, I need to synchronize those threads if and only if the keys of the data object (passed on to the thread) is the same.
My thought process so far:
I tried to have a class that provides the lock on the fly for a given key which looks something like this.
/*
* Singleton class
*/
public class DataAccessKeyToLockProvider {
private volatile Map<String, ReentrantLock> accessKeyToLockHolder = new ConcurrentHashMap<>();
private DataAccessKeyToLockProvider() {
}
public ReentrantLock getLock(String key) {
return accessKeyToLockHolder.putIfAbsent(key, new ReentrantLock());
}
public void removeLock(BSSKey key) {
ReentrantLock removedLock = accessKeyToLockHolder.remove(key);
}
}
So each thread would call this class and get the lock and use it and remove it once the processing is done. But this can so result in a case where the second thread could get the lock object that was inserted by the first thread and waiting for the first thread to release the lock. Once the first thread removes the lock, now the third thread would get a different lock altogether, so the 2nd thread and the 3rd thread are not in sync anymore.
Something like this:
new Thread(() -> {
ReentrantLock lock = DataAccessKeyToLockProvider.get(data.getDataAccessKey());
lock.lock();
data.addData();
lock.unlock();
DataAccessKeyToLockProvider.remove(data.getDataAccessKey());
).start();
Please let me know if you need any additional details to help me resolve my problem
P.S: Removing the key from the lock provider is kind of mandatory as i will be dealing with some millions of keys (not necessarily strings), so I don't want the lock provider to eat up my memory
Inspired the solution provided #rzwitserloot, I have tried to put some generic code that waits for the other thread to complete its processing before giving the access to the next thread.
public class GenericKeyToLockProvider<K> {
private volatile Map<K, ReentrantLock> keyToLockHolder = new ConcurrentHashMap<>();
public synchronized ReentrantLock getLock(K key) {
ReentrantLock existingLock = keyToLockHolder.get(key);
try {
if (existingLock != null && existingLock.isLocked()) {
existingLock.lock(); // Waits for the thread that acquired the lock previously to release it
}
return keyToLockHolder.put(key, new ReentrantLock()); // Override with the new lock
} finally {
if (existingLock != null) {
existingLock.unlock();
}
}
}
}
But looks like the entry made by the last thread wouldn't be removed. Anyway to solve this?
First, a clarification: You either use ReentrantLock, OR you use synchronized. You don't synchronized on a ReentrantLock instance (you synchronize on any object you want) – or, if you want to go the lock route, you can call the lock lock method on your lock object, using a try/finally guard to always ensure you call unlock later (and don't use synchronized at all).
synchronized is low-level API. Lock, and all the other classes in the java.util.concurrent package are higher level and offer far more abstractions. It's generally a good idea to just peruse the javadoc of all the classes in the j.u.c package from time to time, very useful stuff in there.
The key issue is to remove all references to a lock object (thus ensuring it can be garbage collected), but not until you are certain there are zero active threads locking on it. Your current approach does not know how many classes are waiting. That needs to be fixed. Once you return an instance of a Lock object, it's 'out of your hands' and it is not possible to track if the caller is ever going to call lock on it. Thus, you can't do that. Instead, call lock as part of the job; the getLock method should actually do the locking as part of the operation. That way, YOU get to control the process flow. However, let's first take a step back:
You say you'll have millions of keys. Okay; but it is somewhat unlikely you'll have millions of threads. After all, a thread requires a stack, and even using the -Xss parameter to reduce the stack size to the minimum of 128k or so, a million threads implies you're using up 128GB of RAM just for stacks; seems unlikely.
So, whilst you might have millions of keys, the number of 'locked' keys is MUCH smaller. Let's focus on those.
You could make a ConcurrentHashMap which maps your string keys to lock objects. Then:
To acquire a lock:
Create a new lock object (literally: Object o = new Object(); - we are going to be using synchronized) and add it to the map using putIfAbsent. If you managed to create the key/value pair (compare the returned object using == to the one you made; if they are the same, you were the one to add it), you got it, go, run the code. Once you're done, acquire the sync lock on your object, send a notification, release, and remove:
public void doWithLocking(String key, Runnable op) {
Object locker = new Object();
Object o = concurrentMap.putIfAbsent(key, locker);
if (o == locker) {
op.run();
synchronized (locker) {
locker.notifyAll(); // wake up everybody waiting.
concurrentMap.remove(key); // this has to be inside!
}
} else {
...
}
}
To wait until the lock is available, first acquire a lock on the locker object, THEN check if the concurrentMap still contains it. If not, you're now free to retry this operation. If it's still in, then we now wait for a notification. In any case we always just retry from scratch. Thus:
public void performWithLocking(String key, Runnable op) throws InterruptedException {
while (true) {
Object locker = new Object();
Object o = concurrentMap.putIfAbsent(key, locker);
if (o == locker) {
try {
op.run();
} finally {
// We want to lock even if the operation throws!
synchronized (locker) {
locker.notifyAll(); // wake up everybody waiting.
concurrentMap.remove(key); // this has to be inside!
}
}
return;
} else {
synchronized (o) {
if (concurrentMap.containsKey(key)) o.wait();
}
}
}
}
}
Instead of this setup where you pass the operation to execute along with the lock key, you could have tandem 'lock' and 'unlock' methods but now you run the risk of writing code that forgets to call unlock. Hence why I wouldn't advise it!
You can call this with, for example:
keyedLockSupportThingie.doWithLocking("mykey", () -> {
System.out.println("Hello, from safety!");
});
My problem is a little complicated:
I have a concurrent map, the threads want to visit and update the map, if two thread want to fetch the same entry of the map, one should first get the map, update it, and the other should wait until the entry has been updated successfully and then fetch the entry.
My original thought is that: I can use another concurrent map, same key with the targeting map and use a latch as its value.
My code is like:
private final ConcurrentMap<Long, List<RowKeyMap>> targetmap;
private final ConcurrentMap<Long, CountDownLatch> helpermap;
long keyMillis; //key
CountDownLatch restoreLatch = helpermap.get(keyMillis);
if (restoreLatch != null) {
try {
restoreLatch.await();
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
throw new RuntimeException("Interrupted trying to get " + keyMillis);
}
}
List<RowKeyMap> restoredata = targetmap.get(keyMillis);
if (restoredata == null) {
//find the entry should be restored, put a latch into the helpermap and restore it
restoreLatch = new CountDownLatch(1);
CountDownLatch existingLatch = helpermap.putIfAbsent(keyMillis, restoreLatch);
if (existingLatch == null) {
microshards = new ArrayList<>(count);
for (int i = 0; i < count; ++i) {
microshards.add(new RowKeyMap(some parameters));
}
List<RowKeyMap> existing = targetmap.putIfAbsent(keyMillis, microshards);
if (existing == null) {
{do actual restore job here}
} else {
microshards = existing;
}
restoreLatch.countDown();
restoresByDate.remove(keyMillis);
} else {
// Lost the race, wait for the restore task is complete and get the new restoredata
try {
existingLatch.await();
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
throw new RuntimeException("Interrupted trying to get " + keyMillis);
}
{get the new restoredata}
}
}
But the current version has a bug:
Thread A executes through first line, gets null for restoreLatch
Thread B wakes up and executes through first line, also gets null for
restoreLatch
Thread B continues on to following lines, sees existingLatch is null
Thread B continues, puts a created-but-not-yet-restored-into list
into restoredata
Thread A wakes up and executes through, get the
created-but-not-yet-restored-into list from restoredata
Anyone has any ideas to solve this? Thanks!
So what you want is a lock per each map entry. I'm not sure a CountDownLatch is ideal here, because it cannot be re-used, and creating a new one each time complicates your problem.
But your basic problem is that you are not preventing the race condition for the lock itself.
In order to do that, you must first ensure that a lock object for the entry exists, and that if two threads go to the same entry, they will get the same lock.
You can do this by first creating a lock object, then use putIfAbsent to put it in the lock map:
Object entryLock = new Object();
Object returnedLock = helpermap.putIfAbsent( keyMillis, entryLock );
entryLock = returnedLock == null ? entryLock : returnedLock;
What this does is ensure that any two threads that are trying to access the same entry (keyMillis) will get the same lock instance. If thread A is the first to run the putIfAbsent line, then the new object it created in the first line is going to be the one to be placed in the helper map, and it's going to get null back, which means it will also use the object it just placed in the map - entryLock. Thread B then comes along and create its own entryLock. But when it tries the putIfAbsent line, there is already an object mapped to keyMillis, returnedLock, and that's the object it will use (in this case, the original new lock it created will be discarded to the garbage collection).
So whichever order thy get to the putIfAbsent line, they will be using the same lock instance. Now the next step is:
Lock the lock.
Run your processing of the data in the targetMap, creating it if it doesn't exist, updating it if it does, etc. All this time, other threads for this particular keyMillis are waiting, but threads with other keyMillis don't.
Unlock the lock. One of the other threads that wait for this keyMillis entry will now lock the lock.
To do this is pretty simple:
synchronized(entryLock) {
// All operations on the particular entry
}
If you need fancier lock facilities then use ReentrantLock or a CyclicBarrier. A CountDownLatch will need to be replaced with a new one to be usable, and that would defeat the arrangement above, which pretty much relies on the lock object being the same for all threads.
My problem is a little complicated...
Uh oh, I've got bad news for you. If you think that problem is complicated.... You actually have identified the simplest problem in multi-threaded program. It's called mutual exclusion.
You can do it like this in Java:
final Object lock = new Object();
synchronized (lock) {
// some protected statements
}
synchronized (lock) {
// some more protected statements
}
The JVM guarantees that no more than one thread can be in a synchronized(foo) block for the same object foo at the same time.
use the synchronized either at the method declaration or inside a block of code.
see here : https://docs.oracle.com/javase/tutorial/essential/concurrency/syncmeth.html
Use objects that implement the Lock interface. Java has several of these objects by default that you can use. Here is a tutorial that explains what locks are and how to use them.
I would like to collect some metrics from various places in a web app. To keep it simple, all these will be counters and therefore the only modifier operation is to increment them by 1.
The increments will be concurrent and often. The reads (dumping the stats) is a rare operation.
I was thinking to use a ConcurrentHashMap. The issue is how to increment the counters correctly. Since the map doesn't have an "increment" operation, I need to read the current value first, increment it than put the new value in the map. Without more code, this is not an atomic operation.
Is it possible to achieve this without synchronization (which would defeat the purpose of the ConcurrentHashMap)? Do I need to look at Guava ?
Thanks for any pointers.
P.S.
There is a related question on SO (Most efficient way to increment a Map value in Java) but focused on performance and not multi-threading
UPDATE
For those arriving here through searches on the same topic: besides the answers below, there's a useful presentation which incidentally covers the same topic. See slides 24-33.
In Java 8:
ConcurrentHashMap<String, LongAdder> map = new ConcurrentHashMap<>();
map.computeIfAbsent("key", k -> new LongAdder()).increment();
Guava's new AtomicLongMap (in release 11) might address this need.
You're pretty close. Why don't you try something like a ConcurrentHashMap<Key, AtomicLong>?
If your Keys (metrics) are unchanging, you could even just use a standard HashMap (they are threadsafe if readonly, but you'd be well advised to make this explicit with an ImmutableMap from Google Collections or Collections.unmodifiableMap, etc.).
This way, you can use map.get(myKey).incrementAndGet() to bump statistics.
Other than going with AtomicLong, you can do the usual cas-loop thing:
private final ConcurrentMap<Key,Long> counts =
new ConcurrentHashMap<Key,Long>();
public void increment(Key key) {
if (counts.putIfAbsent(key, 1)) == null) {
return;
}
Long old;
do {
old = counts.get(key);
} while (!counts.replace(key, old, old+1)); // Assumes no removal.
}
(I've not written a do-while loop for ages.)
For small values the Long will probably be "cached". For longer values, it may require allocation. But the allocations are actually extremely fast (and you can cache further) - depends upon what you expect, in the worst case.
Got a necessity to do the same.
I'm using ConcurrentHashMap + AtomicInteger.
Also, ReentrantRW Lock was introduced for atomic flush(very similar behavior).
Tested with 10 Keys and 10 Threads per each Key. Nothing was lost.
I just haven't tried several flushing threads yet, but hope it will work.
Massive singleusermode flush is torturing me...
I want to remove RWLock and break down flushing into small pieces. Tomorrow.
private ConcurrentHashMap<String,AtomicInteger> counters = new ConcurrentHashMap<String, AtomicInteger>();
private ReadWriteLock rwLock = new ReentrantReadWriteLock();
public void count(String invoker) {
rwLock.readLock().lock();
try{
AtomicInteger currentValue = counters.get(invoker);
// if entry is absent - initialize it. If other thread has added value before - we will yield and not replace existing value
if(currentValue == null){
// value we want to init with
AtomicInteger newValue = new AtomicInteger(0);
// try to put and get old
AtomicInteger oldValue = counters.putIfAbsent(invoker, newValue);
// if old value not null - our insertion failed, lets use old value as it's in the map
// if old value is null - our value was inserted - lets use it
currentValue = oldValue != null ? oldValue : newValue;
}
// counter +1
currentValue.incrementAndGet();
}finally {
rwLock.readLock().unlock();
}
}
/**
* #return Map with counting results
*/
public Map<String, Integer> getCount() {
// stop all updates (readlocks)
rwLock.writeLock().lock();
try{
HashMap<String, Integer> resultMap = new HashMap<String, Integer>();
// read all Integers to a new map
for(Map.Entry<String,AtomicInteger> entry: counters.entrySet()){
resultMap.put(entry.getKey(), entry.getValue().intValue());
}
// reset ConcurrentMap
counters.clear();
return resultMap;
}finally {
rwLock.writeLock().unlock();
}
}
I did a benchmark to compare the performance of LongAdder and AtomicLong.
LongAdder had a better performance in my benchmark: for 500 iterations using a map with size 100 (10 concurrent threads), the average time for LongAdder was 1270ms while that for AtomicLong was 1315ms.
I have a webapp that I am in the middle of doing some load/performance testing on, particularily on a feature where we expect a few hundred users to be accessing the same page and hitting refresh about every 10 seconds on this page. One area of improvement that we found we could make with this function was to cache the responses from the web service for some period of time, since the data is not changing.
After implementing this basic caching, in some further testing I found out that I didn't consider how concurrent threads could access the Cache at the same time. I found that within the matter of ~100ms, about 50 threads were trying to fetch the object from the Cache, finding that it had expired, hitting the web service to fetch the data, and then putting the object back in the cache.
The original code looked something like this:
private SomeData[] getSomeDataByEmail(WebServiceInterface service, String email) {
final String key = "Data-" + email;
SomeData[] data = (SomeData[]) StaticCache.get(key);
if (data == null) {
data = service.getSomeDataForEmail(email);
StaticCache.set(key, data, CACHE_TIME);
}
else {
logger.debug("getSomeDataForEmail: using cached object");
}
return data;
}
So, to make sure that only one thread was calling the web service when the object at key expired, I thought I needed to synchronize the Cache get/set operation, and it seemed like using the cache key would be a good candidate for an object to synchronize on (this way, calls to this method for email b#b.com would not be blocked by method calls to a#a.com).
I updated the method to look like this:
private SomeData[] getSomeDataByEmail(WebServiceInterface service, String email) {
SomeData[] data = null;
final String key = "Data-" + email;
synchronized(key) {
data =(SomeData[]) StaticCache.get(key);
if (data == null) {
data = service.getSomeDataForEmail(email);
StaticCache.set(key, data, CACHE_TIME);
}
else {
logger.debug("getSomeDataForEmail: using cached object");
}
}
return data;
}
I also added logging lines for things like "before synchronization block", "inside synchronization block", "about to leave synchronization block", and "after synchronization block", so I could determine if I was effectively synchronizing the get/set operation.
However it doesn't seem like this has worked. My test logs have output like:
(log output is 'threadname' 'logger name' 'message')
http-80-Processor253 jsp.view-page - getSomeDataForEmail: about to enter synchronization block
http-80-Processor253 jsp.view-page - getSomeDataForEmail: inside synchronization block
http-80-Processor253 cache.StaticCache - get: object at key [SomeData-test#test.com] has expired
http-80-Processor253 cache.StaticCache - get: key [SomeData-test#test.com] returning value [null]
http-80-Processor263 jsp.view-page - getSomeDataForEmail: about to enter synchronization block
http-80-Processor263 jsp.view-page - getSomeDataForEmail: inside synchronization block
http-80-Processor263 cache.StaticCache - get: object at key [SomeData-test#test.com] has expired
http-80-Processor263 cache.StaticCache - get: key [SomeData-test#test.com] returning value [null]
http-80-Processor131 jsp.view-page - getSomeDataForEmail: about to enter synchronization block
http-80-Processor131 jsp.view-page - getSomeDataForEmail: inside synchronization block
http-80-Processor131 cache.StaticCache - get: object at key [SomeData-test#test.com] has expired
http-80-Processor131 cache.StaticCache - get: key [SomeData-test#test.com] returning value [null]
http-80-Processor104 jsp.view-page - getSomeDataForEmail: inside synchronization block
http-80-Processor104 cache.StaticCache - get: object at key [SomeData-test#test.com] has expired
http-80-Processor104 cache.StaticCache - get: key [SomeData-test#test.com] returning value [null]
http-80-Processor252 jsp.view-page - getSomeDataForEmail: about to enter synchronization block
http-80-Processor283 jsp.view-page - getSomeDataForEmail: about to enter synchronization block
http-80-Processor2 jsp.view-page - getSomeDataForEmail: about to enter synchronization block
http-80-Processor2 jsp.view-page - getSomeDataForEmail: inside synchronization block
I wanted to see only one thread at a time entering/exiting the synchronization block around the get/set operations.
Is there an issue in synchronizing on String objects? I thought the cache-key would be a good choice as it is unique to the operation, and even though the final String key is declared within the method, I was thinking that each thread would be getting a reference to the same object and therefore would synchronization on this single object.
What am I doing wrong here?
Update: after looking further at the logs, it seems like methods with the same synchronization logic where the key is always the same, such as
final String key = "blah";
...
synchronized(key) { ...
do not exhibit the same concurrency problem - only one thread at a time is entering the block.
Update 2: Thanks to everyone for the help! I accepted the first answer about intern()ing Strings, which solved my initial problem - where multiple threads were entering synchronized blocks where I thought they shouldn't, because the key's had the same value.
As others have pointed out, using intern() for such a purpose and synchronizing on those Strings does indeed turn out to be a bad idea - when running JMeter tests against the webapp to simulate the expected load, I saw the used heap size grow to almost 1GB in just under 20 minutes.
Currently I'm using the simple solution of just synchronizing the entire method - but I really like the code samples provided by martinprobst and MBCook, but since I have about 7 similar getData() methods in this class currently (since it needs about 7 different pieces of data from a web service), I didn't want to add almost-duplicate logic about getting and releasing locks to each method. But this is definitely very, very valuable info for future usage. I think these are ultimately the correct answers on how best to make an operation like this thread-safe, and I'd give out more votes to these answers if I could!
Without putting my brain fully into gear, from a quick scan of what you say it looks as though you need to intern() your Strings:
final String firstkey = "Data-" + email;
final String key = firstkey.intern();
Two Strings with the same value are otherwise not necessarily the same object.
Note that this may introduce a new point of contention, since deep in the VM, intern() may have to acquire a lock. I have no idea what modern VMs look like in this area, but one hopes they are fiendishly optimised.
I assume you know that StaticCache still needs to be thread-safe. But the contention there should be tiny compared with what you'd have if you were locking on the cache rather than just the key while calling getSomeDataForEmail.
Response to question update:
I think that's because a string literal always yields the same object. Dave Costa points out in a comment that it's even better than that: a literal always yields the canonical representation. So all String literals with the same value anywhere in the program would yield the same object.
Edit
Others have pointed out that synchronizing on intern strings is actually a really bad idea - partly because creating intern strings is permitted to cause them to exist in perpetuity, and partly because if more than one bit of code anywhere in your program synchronizes on intern strings, you have dependencies between those bits of code, and preventing deadlocks or other bugs may be impossible.
Strategies to avoid this by storing a lock object per key string are being developed in other answers as I type.
Here's an alternative - it still uses a singular lock, but we know we're going to need one of those for the cache anyway, and you were talking about 50 threads, not 5000, so that may not be fatal. I'm also assuming that the performance bottleneck here is slow blocking I/O in DoSlowThing() which will therefore hugely benefit from not being serialised. If that's not the bottleneck, then:
If the CPU is busy then this approach may not be sufficient and you need another approach.
If the CPU is not busy, and access to server is not a bottleneck, then this approach is overkill, and you might as well forget both this and per-key locking, put a big synchronized(StaticCache) around the whole operation, and do it the easy way.
Obviously this approach needs to be soak tested for scalability before use -- I guarantee nothing.
This code does NOT require that StaticCache is synchronized or otherwise thread-safe. That needs to be revisited if any other code (for example scheduled clean-up of old data) ever touches the cache.
IN_PROGRESS is a dummy value - not exactly clean, but the code's simple and it saves having two hashtables. It doesn't handle InterruptedException because I don't know what your app wants to do in that case. Also, if DoSlowThing() consistently fails for a given key this code as it stands is not exactly elegant, since every thread through will retry it. Since I don't know what the failure criteria are, and whether they are liable to be temporary or permanent, I don't handle this either, I just make sure threads don't block forever. In practice you may want to put a data value in the cache which indicates 'not available', perhaps with a reason, and a timeout for when to retry.
// do not attempt double-check locking here. I mean it.
synchronized(StaticObject) {
data = StaticCache.get(key);
while (data == IN_PROGRESS) {
// another thread is getting the data
StaticObject.wait();
data = StaticCache.get(key);
}
if (data == null) {
// we must get the data
StaticCache.put(key, IN_PROGRESS, TIME_MAX_VALUE);
}
}
if (data == null) {
// we must get the data
try {
data = server.DoSlowThing(key);
} finally {
synchronized(StaticObject) {
// WARNING: failure here is fatal, and must be allowed to terminate
// the app or else waiters will be left forever. Choose a suitable
// collection type in which replacing the value for a key is guaranteed.
StaticCache.put(key, data, CURRENT_TIME);
StaticObject.notifyAll();
}
}
}
Every time anything is added to the cache, all threads wake up and check the cache (no matter what key they're after), so it's possible to get better performance with less contentious algorithms. However, much of that work will take place during your copious idle CPU time blocking on I/O, so it may not be a problem.
This code could be commoned-up for use with multiple caches, if you define suitable abstractions for the cache and its associated lock, the data it returns, the IN_PROGRESS dummy, and the slow operation to perform. Rolling the whole thing into a method on the cache might not be a bad idea.
Synchronizing on an intern'd String might not be a good idea at all - by interning it, the String turns into a global object, and if you synchronize on the same interned strings in different parts of your application, you might get really weird and basically undebuggable synchronization issues such as deadlocks. It might seem unlikely, but when it happens you are really screwed. As a general rule, only ever synchronize on a local object where you're absolutely sure that no code outside of your module might lock it.
In your case, you can use a synchronized hashtable to store locking objects for your keys.
E.g.:
Object data = StaticCache.get(key, ...);
if (data == null) {
Object lock = lockTable.get(key);
if (lock == null) {
// we're the only one looking for this
lock = new Object();
synchronized(lock) {
lockTable.put(key, lock);
// get stuff
lockTable.remove(key);
}
} else {
synchronized(lock) {
// just to wait for the updater
}
data = StaticCache.get(key);
}
} else {
// use from cache
}
This code has a race condition, where two threads might put an object into the lock table after each other. This should however not be a problem, because then you only have one more thread calling the webservice and updating the cache, which shouldn't be a problem.
If you're invalidating the cache after some time, you should check whether data is null again after retrieving it from the cache, in the lock != null case.
Alternatively, and much easier, you can make the whole cache lookup method ("getSomeDataByEmail") synchronized. This will mean that all threads have to synchronize when they access the cache, which might be a performance problem. But as always, try this simple solution first and see if it's really a problem! In many cases it should not be, as you probably spend much more time processing the result than synchronizing.
Strings are not good candidates for synchronization. If you must synchronize on a String ID, it can be done by using the string to create a mutex (see "synchronizing on an ID"). Whether the cost of that algorithm is worth it depends on whether invoking your service involves any significant I/O.
Also:
I hope the StaticCache.get() and set() methods are threadsafe.
String.intern() comes at a cost (one that varies between VM implementations) and should be used with care.
Here is a safe short Java 8 solution that uses a map of dedicated lock objects for synchronization:
private static final Map<String, Object> keyLocks = new ConcurrentHashMap<>();
private SomeData[] getSomeDataByEmail(WebServiceInterface service, String email) {
final String key = "Data-" + email;
synchronized (keyLocks.computeIfAbsent(key, k -> new Object())) {
SomeData[] data = StaticCache.get(key);
if (data == null) {
data = service.getSomeDataForEmail(email);
StaticCache.set(key, data);
}
}
return data;
}
It has a drawback that keys and lock objects would retain in map forever.
This can be worked around like this:
private SomeData[] getSomeDataByEmail(WebServiceInterface service, String email) {
final String key = "Data-" + email;
synchronized (keyLocks.computeIfAbsent(key, k -> new Object())) {
try {
SomeData[] data = StaticCache.get(key);
if (data == null) {
data = service.getSomeDataForEmail(email);
StaticCache.set(key, data);
}
} finally {
keyLocks.remove(key); // vulnerable to race-conditions
}
}
return data;
}
But then popular keys would be constantly reinserted in map with lock objects being reallocated.
Update: And this leaves race condition possibility when two threads would concurrently enter synchronized section for the same key but with different locks.
So it may be more safe and efficient to use expiring Guava Cache:
private static final LoadingCache<String, Object> keyLocks = CacheBuilder.newBuilder()
.expireAfterAccess(10, TimeUnit.MINUTES) // max lock time ever expected
.build(CacheLoader.from(Object::new));
private SomeData[] getSomeDataByEmail(WebServiceInterface service, String email) {
final String key = "Data-" + email;
synchronized (keyLocks.getUnchecked(key)) {
SomeData[] data = StaticCache.get(key);
if (data == null) {
data = service.getSomeDataForEmail(email);
StaticCache.set(key, data);
}
}
return data;
}
Note that it's assumed here that StaticCache is thread-safe and wouldn't suffer from concurrent reads and writes for different keys.
Others have suggested interning the strings, and that will work.
The problem is that Java has to keep interned strings around. I was told it does this even if you're not holding a reference because the value needs to be the same the next time someone uses that string. This means interning all the strings may start eating up memory, which with the load you're describing could be a big problem.
I have seen two solutions to this:
You could synchronize on another object
Instead of the email, make an object that holds the email (say the User object) that holds the value of email as a variable. If you already have another object that represents the person (say you already pulled something from the DB based on their email) you could use that. By implementing the equals method and the hashcode method you can make sure Java considers the objects the same when you do a static cache.contains() to find out if the data is already in the cache (you'll have to synchronize on the cache).
Actually, you could keep a second Map for objects to lock on. Something like this:
Map<String, Object> emailLocks = new HashMap<String, Object>();
Object lock = null;
synchronized (emailLocks) {
lock = emailLocks.get(emailAddress);
if (lock == null) {
lock = new Object();
emailLocks.put(emailAddress, lock);
}
}
synchronized (lock) {
// See if this email is in the cache
// If so, serve that
// If not, generate the data
// Since each of this person's threads synchronizes on this, they won't run
// over eachother. Since this lock is only for this person, it won't effect
// other people. The other synchronized block (on emailLocks) is small enough
// it shouldn't cause a performance problem.
}
This will prevent 15 fetches on the same email address at one. You'll need something to prevent too many entries from ending up in the emailLocks map. Using LRUMaps from Apache Commons would do it.
This will need some tweaking, but it may solve your problem.
Use a different key
If you are willing to put up with possible errors (I don't know how important this is) you could use the hashcode of the String as the key. ints don't need to be interned.
Summary
I hope this helps. Threading is fun, isn't it? You could also use the session to set a value meaning "I'm already working on finding this" and check that to see if the second (third, Nth) thread needs to attempt to create the or just wait for the result to show up in the cache. I guess I had three suggestions.
You can use the 1.5 concurrency utilities to provide a cache designed to allow multiple concurrent access, and a single point of addition (i.e. only one thread ever performing the expensive object "creation"):
private ConcurrentMap<String, Future<SomeData[]> cache;
private SomeData[] getSomeDataByEmail(final WebServiceInterface service, final String email) throws Exception {
final String key = "Data-" + email;
Callable<SomeData[]> call = new Callable<SomeData[]>() {
public SomeData[] call() {
return service.getSomeDataForEmail(email);
}
}
FutureTask<SomeData[]> ft; ;
Future<SomeData[]> f = cache.putIfAbsent(key, ft= new FutureTask<SomeData[]>(call)); //atomic
if (f == null) { //this means that the cache had no mapping for the key
f = ft;
ft.run();
}
return f.get(); //wait on the result being available if it is being calculated in another thread
}
Obviously, this doesn't handle exceptions as you'd want to, and the cache doesn't have eviction built in. Perhaps you could use it as a basis to change your StaticCache class, though.
Use a decent caching framework such as ehcache.
Implementing a good cache is not as easy as some people believe.
Regarding the comment that String.intern() is a source of memory leaks, that is actually not true.
Interned Strings are garbage collected,it just might take longer because on certain JVM'S (SUN) they are stored in Perm space which is only touched by full GC's.
Your main problem is not just that there might be multiple instances of String with the same value. The main problem is that you need to have only one monitor on which to synchronize for accessing the StaticCache object. Otherwise multiple threads might end up concurrently modifying StaticCache (albeit under different keys), which most likely doesn't support concurrent modification.
The call:
final String key = "Data-" + email;
creates a new object every time the method is called. Because that object is what you use to lock, and every call to this method creates a new object, then you are not really synchronizing access to the map based on the key.
This further explain your edit. When you have a static string, then it will work.
Using intern() solves the problem, because it returns the string from an internal pool kept by the String class, that ensures that if two strings are equal, the one in the pool will be used. See
http://java.sun.com/j2se/1.4.2/docs/api/java/lang/String.html#intern()
This question seems to me a bit too broad, and therefore it instigated equally broad set of answers. So I'll try to answer the question I have been redirected from, unfortunately that one has been closed as duplicate.
public class ValueLock<T> {
private Lock lock = new ReentrantLock();
private Map<T, Condition> conditions = new HashMap<T, Condition>();
public void lock(T t){
lock.lock();
try {
while (conditions.containsKey(t)){
conditions.get(t).awaitUninterruptibly();
}
conditions.put(t, lock.newCondition());
} finally {
lock.unlock();
}
}
public void unlock(T t){
lock.lock();
try {
Condition condition = conditions.get(t);
if (condition == null)
throw new IllegalStateException();// possibly an attempt to release what wasn't acquired
conditions.remove(t);
condition.signalAll();
} finally {
lock.unlock();
}
}
Upon the (outer) lock operation the (inner) lock is acquired to get an exclusive access to the map for a short time, and if the correspondent object is already in the map, the current thread will wait,
otherwise it will put new Condition to the map, release the (inner) lock and proceed,
and the (outer) lock is considered obtained.
The (outer) unlock operation, first acquiring an (inner) lock, will signal on Condition and then remove the object from the map.
The class does not use concurrent version of Map, because every access to it is guarded by single (inner) lock.
Please notice, the semantic of lock() method of this class is different that of ReentrantLock.lock(), the repeated lock() invocations without paired unlock() will hang current thread indefinitely.
An example of usage that might be applicable to the situation, the OP described
ValueLock<String> lock = new ValueLock<String>();
// ... share the lock
String email = "...";
try {
lock.lock(email);
//...
} finally {
lock.unlock(email);
}
This is rather late, but there is quite a lot of incorrect code presented here.
In this example:
private SomeData[] getSomeDataByEmail(WebServiceInterface service, String email) {
SomeData[] data = null;
final String key = "Data-" + email;
synchronized(key) {
data =(SomeData[]) StaticCache.get(key);
if (data == null) {
data = service.getSomeDataForEmail(email);
StaticCache.set(key, data, CACHE_TIME);
}
else {
logger.debug("getSomeDataForEmail: using cached object");
}
}
return data;
}
The synchronization is incorrectly scoped. For a static cache that supports a get/put API, there should be at least synchronization around the get and getIfAbsentPut type operations, for safe access to the cache. The scope of synchronization will be the cache itself.
If updates must be made to the data elements themselves, that adds an additional layer of synchronization, which should be on the individual data elements.
SynchronizedMap can be used in place of explicit synchronization, but care must still be observed. If the wrong APIs are used (get and put instead of putIfAbsent) then the operations won't have the necessary synchronization, despite the use of the synchronized map. Notice the complications introduced by the use of putIfAbsent: Either, the put value must be computed even in cases when it is not needed (because the put cannot know if the put value is needed until the cache contents are examined), or requires a careful use of delegation (say, using Future, which works, but is somewhat of a mismatch; see below), where the put value is obtained on demand if needed.
The use of Futures is possible, but seems rather awkward, and perhaps a bit of overengineering. The Future API is at it's core for asynchronous operations, in particular, for operations which may not complete immediately. Involving Future very probably adds a layer of thread creation -- extra probably unnecessary complications.
The main problem of using Future for this type of operation is that Future inherently ties in multi-threading. Use of Future when a new thread is not necessary means ignoring a lot of the machinery of Future, making it an overly heavy API for this use.
Latest update 2019,
If you are searching for new ways of implementing synchronization in JAVA, this answer is for you.
I found this amazing blog by Anatoliy Korovin this will help you understand the syncronized deeply.
How to Synchronize Blocks by the Value of the Object in Java.
This helped me hope new developers will find this useful too.
Why not just render a static html page that gets served to the user and regenerated every x minutes?
I'd also suggest getting rid of the string concatenation entirely if you don't need it.
final String key = "Data-" + email;
Is there other things/types of objects in the cache that use the email address that you need that extra "Data-" at the beginning of the key?
if not, i'd just make that
final String key = email;
and you avoid all that extra string creation too.
In case others have a similar problem, the following code works, as far as I can tell:
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.atomic.AtomicInteger;
import java.util.function.Supplier;
public class KeySynchronizer<T> {
private Map<T, CounterLock> locks = new ConcurrentHashMap<>();
public <U> U synchronize(T key, Supplier<U> supplier) {
CounterLock lock = locks.compute(key, (k, v) ->
v == null ? new CounterLock() : v.increment());
synchronized (lock) {
try {
return supplier.get();
} finally {
if (lock.decrement() == 0) {
// Only removes if key still points to the same value,
// to avoid issue described below.
locks.remove(key, lock);
}
}
}
}
private static final class CounterLock {
private AtomicInteger remaining = new AtomicInteger(1);
private CounterLock increment() {
// Returning a new CounterLock object if remaining = 0 to ensure that
// the lock is not removed in step 5 of the following execution sequence:
// 1) Thread 1 obtains a new CounterLock object from locks.compute (after evaluating "v == null" to true)
// 2) Thread 2 evaluates "v == null" to false in locks.compute
// 3) Thread 1 calls lock.decrement() which sets remaining = 0
// 4) Thread 2 calls v.increment() in locks.compute
// 5) Thread 1 calls locks.remove(key, lock)
return remaining.getAndIncrement() == 0 ? new CounterLock() : this;
}
private int decrement() {
return remaining.decrementAndGet();
}
}
}
In the case of the OP, it would be used like this:
private KeySynchronizer<String> keySynchronizer = new KeySynchronizer<>();
private SomeData[] getSomeDataByEmail(WebServiceInterface service, String email) {
String key = "Data-" + email;
return keySynchronizer.synchronize(key, () -> {
SomeData[] existing = (SomeData[]) StaticCache.get(key);
if (existing == null) {
SomeData[] data = service.getSomeDataForEmail(email);
StaticCache.set(key, data, CACHE_TIME);
return data;
}
logger.debug("getSomeDataForEmail: using cached object");
return existing;
});
}
If nothing should be returned from the synchronized code, the synchronize method can be written like this:
public void synchronize(T key, Runnable runnable) {
CounterLock lock = locks.compute(key, (k, v) ->
v == null ? new CounterLock() : v.increment());
synchronized (lock) {
try {
runnable.run();
} finally {
if (lock.decrement() == 0) {
// Only removes if key still points to the same value,
// to avoid issue described below.
locks.remove(key, lock);
}
}
}
}
I've added a small lock class that can lock/synchronize on any key, including strings.
See implementation for Java 8, Java 6 and a small test.
Java 8:
public class DynamicKeyLock<T> implements Lock
{
private final static ConcurrentHashMap<Object, LockAndCounter> locksMap = new ConcurrentHashMap<>();
private final T key;
public DynamicKeyLock(T lockKey)
{
this.key = lockKey;
}
private static class LockAndCounter
{
private final Lock lock = new ReentrantLock();
private final AtomicInteger counter = new AtomicInteger(0);
}
private LockAndCounter getLock()
{
return locksMap.compute(key, (key, lockAndCounterInner) ->
{
if (lockAndCounterInner == null) {
lockAndCounterInner = new LockAndCounter();
}
lockAndCounterInner.counter.incrementAndGet();
return lockAndCounterInner;
});
}
private void cleanupLock(LockAndCounter lockAndCounterOuter)
{
if (lockAndCounterOuter.counter.decrementAndGet() == 0)
{
locksMap.compute(key, (key, lockAndCounterInner) ->
{
if (lockAndCounterInner == null || lockAndCounterInner.counter.get() == 0) {
return null;
}
return lockAndCounterInner;
});
}
}
#Override
public void lock()
{
LockAndCounter lockAndCounter = getLock();
lockAndCounter.lock.lock();
}
#Override
public void unlock()
{
LockAndCounter lockAndCounter = locksMap.get(key);
lockAndCounter.lock.unlock();
cleanupLock(lockAndCounter);
}
#Override
public void lockInterruptibly() throws InterruptedException
{
LockAndCounter lockAndCounter = getLock();
try
{
lockAndCounter.lock.lockInterruptibly();
}
catch (InterruptedException e)
{
cleanupLock(lockAndCounter);
throw e;
}
}
#Override
public boolean tryLock()
{
LockAndCounter lockAndCounter = getLock();
boolean acquired = lockAndCounter.lock.tryLock();
if (!acquired)
{
cleanupLock(lockAndCounter);
}
return acquired;
}
#Override
public boolean tryLock(long time, TimeUnit unit) throws InterruptedException
{
LockAndCounter lockAndCounter = getLock();
boolean acquired;
try
{
acquired = lockAndCounter.lock.tryLock(time, unit);
}
catch (InterruptedException e)
{
cleanupLock(lockAndCounter);
throw e;
}
if (!acquired)
{
cleanupLock(lockAndCounter);
}
return acquired;
}
#Override
public Condition newCondition()
{
LockAndCounter lockAndCounter = locksMap.get(key);
return lockAndCounter.lock.newCondition();
}
}
Java 6:
public class DynamicKeyLock implements Lock
{
private final static ConcurrentHashMap locksMap = new ConcurrentHashMap();
private final T key;
public DynamicKeyLock(T lockKey) {
this.key = lockKey;
}
private static class LockAndCounter {
private final Lock lock = new ReentrantLock();
private final AtomicInteger counter = new AtomicInteger(0);
}
private LockAndCounter getLock()
{
while (true) // Try to init lock
{
LockAndCounter lockAndCounter = locksMap.get(key);
if (lockAndCounter == null)
{
LockAndCounter newLock = new LockAndCounter();
lockAndCounter = locksMap.putIfAbsent(key, newLock);
if (lockAndCounter == null)
{
lockAndCounter = newLock;
}
}
lockAndCounter.counter.incrementAndGet();
synchronized (lockAndCounter)
{
LockAndCounter lastLockAndCounter = locksMap.get(key);
if (lockAndCounter == lastLockAndCounter)
{
return lockAndCounter;
}
// else some other thread beat us to it, thus try again.
}
}
}
private void cleanupLock(LockAndCounter lockAndCounter)
{
if (lockAndCounter.counter.decrementAndGet() == 0)
{
synchronized (lockAndCounter)
{
if (lockAndCounter.counter.get() == 0)
{
locksMap.remove(key);
}
}
}
}
#Override
public void lock()
{
LockAndCounter lockAndCounter = getLock();
lockAndCounter.lock.lock();
}
#Override
public void unlock()
{
LockAndCounter lockAndCounter = locksMap.get(key);
lockAndCounter.lock.unlock();
cleanupLock(lockAndCounter);
}
#Override
public void lockInterruptibly() throws InterruptedException
{
LockAndCounter lockAndCounter = getLock();
try
{
lockAndCounter.lock.lockInterruptibly();
}
catch (InterruptedException e)
{
cleanupLock(lockAndCounter);
throw e;
}
}
#Override
public boolean tryLock()
{
LockAndCounter lockAndCounter = getLock();
boolean acquired = lockAndCounter.lock.tryLock();
if (!acquired)
{
cleanupLock(lockAndCounter);
}
return acquired;
}
#Override
public boolean tryLock(long time, TimeUnit unit) throws InterruptedException
{
LockAndCounter lockAndCounter = getLock();
boolean acquired;
try
{
acquired = lockAndCounter.lock.tryLock(time, unit);
}
catch (InterruptedException e)
{
cleanupLock(lockAndCounter);
throw e;
}
if (!acquired)
{
cleanupLock(lockAndCounter);
}
return acquired;
}
#Override
public Condition newCondition()
{
LockAndCounter lockAndCounter = locksMap.get(key);
return lockAndCounter.lock.newCondition();
}
}
Test:
public class DynamicKeyLockTest
{
#Test
public void testDifferentKeysDontLock() throws InterruptedException
{
DynamicKeyLock<Object> lock = new DynamicKeyLock<>(new Object());
lock.lock();
AtomicBoolean anotherThreadWasExecuted = new AtomicBoolean(false);
try
{
new Thread(() ->
{
DynamicKeyLock<Object> anotherLock = new DynamicKeyLock<>(new Object());
anotherLock.lock();
try
{
anotherThreadWasExecuted.set(true);
}
finally
{
anotherLock.unlock();
}
}).start();
Thread.sleep(100);
}
finally
{
Assert.assertTrue(anotherThreadWasExecuted.get());
lock.unlock();
}
}
#Test
public void testSameKeysLock() throws InterruptedException
{
Object key = new Object();
DynamicKeyLock<Object> lock = new DynamicKeyLock<>(key);
lock.lock();
AtomicBoolean anotherThreadWasExecuted = new AtomicBoolean(false);
try
{
new Thread(() ->
{
DynamicKeyLock<Object> anotherLock = new DynamicKeyLock<>(key);
anotherLock.lock();
try
{
anotherThreadWasExecuted.set(true);
}
finally
{
anotherLock.unlock();
}
}).start();
Thread.sleep(100);
}
finally
{
Assert.assertFalse(anotherThreadWasExecuted.get());
lock.unlock();
}
}
}
In your case you could use something like this (this doesn't leak any memory):
private Synchronizer<String> synchronizer = new Synchronizer();
private SomeData[] getSomeDataByEmail(WebServiceInterface service, String email) {
String key = "Data-" + email;
return synchronizer.synchronizeOn(key, () -> {
SomeData[] data = (SomeData[]) StaticCache.get(key);
if (data == null) {
data = service.getSomeDataForEmail(email);
StaticCache.set(key, data, CACHE_TIME);
} else {
logger.debug("getSomeDataForEmail: using cached object");
}
return data;
});
}
to use it you just add a dependency:
compile 'com.github.matejtymes:javafixes:1.3.0'
You should be very careful using short lived objects with synchronization. Every Java object has an attached monitor and by default this monitor is deflated; however if 2 threads contend on acquiring the monitor, the monitor gets inflated. If the object would be long lived, this isn't a problem. However if the object is short lived, then cleaning up this inflated monitor can be a serious hit on GC times (so higher latencies and reduced throughput). And it can even be tricky to spot on the GC times since it isn't always listed.
If you do want to synchronize, you could use a java.util.concurrent.Lock. Or make use of a manually crafted striped lock and use the hash of the string as an index on that striped lock. This striped lock you keep around so you don't get the GC problems.
So something like this:
static final Object[] locks = newLockArray();
Object lock = locks[hashToIndex(key.hashcode(),locks.length];
synchronized(lock){
....
}
int hashToIndex(int hash, int length) {
if (hash == Integer.MIN_VALUE return 0;
return abs(hash) % length;
}
other way synchronizing on string object :
String cacheKey = ...;
Object obj = cache.get(cacheKey)
if(obj==null){
synchronized (Integer.valueOf(Math.abs(cacheKey.hashCode()) % 127)){
obj = cache.get(cacheKey)
if(obj==null){
//some cal obtain obj value,and put into cache
}
}
}
You can safely use String.intern for synchronize if you can reasonably guarantee that the string value is unique across your system. UUIDS are a good way to approach this. You can associate a UUID with your actual string key, either via a cache, a map, or maybe even store the uuid as a field on your entity object.
#Service
public class MySyncService{
public Map<String, String> lockMap=new HashMap<String, String>();
public void syncMethod(String email) {
String lock = lockMap.get(email);
if(lock==null) {
lock = UUID.randomUUID().toString();
lockMap.put(email, lock);
}
synchronized(lock.intern()) {
//do your sync code here
}
}