redisson-3.17.2

Feature - RScoredSortedSet.replace() method added
Feature - Spring Data Redis 2.7.0 module added
Feature - RPatternTopic.removeAllListenersAsync() method added
Feature - RShardedTopic object added (requires Redis 7.0+)
Feature - allow to specify username and password in redis connection url
Feature - JCache data partitioning with local cache support

Fixed - "Can't add slave" exceptions after fail over in cluster mode
Fixed - "Unable to acquire subscription" error after connection interruption
Fixed - JCache hangs forever when getting value from cache with useScriptCache=true
Fixed - RMap.merge() method hangs if MapLoader specified
Fixed - FairLock thread counter should start from 1 (thanks to @thisiswanghy)

redisson-3.17.1

Feature - transient fields support for LiveObjects to avoid data serialization
Feature - removeAllListenersAsync() method added to RTopic object
Feature - transactionAware setting added to RedissonSpringCacheManager

Improvement - amount of created connections in parallel reduced to 5, for better stability

Fixed - RedissonReactiveClient.getMultilock() method should accept RLockReactive objects
Fixed - RedissonRxClient.getMultilock() method should accept RLockRx objects
Fixed - don't close connection on error response during topology scan
Fixed - SET command should be an idempotent operation
Fixed - MasterSlaveConnectionManager throws ClassCastException if host unknown
Fixed - RReadWriteLock renewal doesn't work if writeLock released before readLock then both were acquired
Fixed - Spring Data Redis module. Scan In cluster mode, other nodes cannot be scanned
Fixed - RReliableTopic object throws "attempt to compare nil with number" error
Fixed - RedissonSpinLock.tryLock() method returns false instead of true if the remaining wait time is negative
Fixed - an error should be thrown if merge(), compute(), computeIfAbsent() and computeIfPresent() of RMap used in batch
Fixed - Unable to specify timezone in CronSchedule object
Fixed - RMapCache.destroy() method throws NPE
Fixed - RLock.tryLock() method throws CancellationException
Fixed - Unable to connect to Redis server error is thrown due to NPE
Fixed - RBlockingQueue.pollLastAndOfferFirstTo() throws ClassCastException if result is empty
Fixed - internal AsyncSemaphore doesn't skip canceled tasks in the same thread
Fixed - RLocalCachedMap.getAll() method doesn't respect storeCacheMiss setting
Fixed - 0 value for waitTime and leastTime isn't handled correctly by RMultiLock object
Fixed - Spring Data Redis module. RedissonConnection.execute() method doesn't invoke overloaded methods correctly

redisson-3.17.0

Feature - RFunction object added (requires Redis 7.0+)
Feature - pollLastEntriesFromAny() and pollFirstEntriesFromAny() methods added to RScoredSortedSet object (requires Redis 7.0+)
Feature - expireIfSet(), expireIfNotSet(), expireIfGreater() and expireIfLess() methods added to RExpirable interface (requires Redis 7.0+)
Feature - checkLockSyncedSlaves setting added
Feature - getAndExpire and getAndClearExpire() methods added to RBucket object (requires Redis 6.2.0+)
Feature - pollFirstFromAny() and pollLastFromAny() methods with timeout and count added to RScoredSortedSet object (requires Redis 7.0+)
Feature - pollFirst() and pollLast() methods with timeout and count added to RScoredSortedSet object (requires Redis 7.0+)
Feature - addAllIfLess(), addAllIfGreater(), addAllIfExist(), addAllIfAbsent() methods added to RScoredSortedSet object
Feature - RExpirable.expire(Duration) method added
Feature - RExpirable.expireTime() method added (requires Redis 7.0+)
Feature - range(), rangeReversed(), entryRange(), entryRangeReversed() methods with limit parameter added to RTimeSeries object
Feature - TransactionalOperation.syncSlaves setting added
Feature - pollFirstFromAny() and pollLastFromAny() methods added to RBlockingQueue object (requires Redis 7.0+)

Improvement - read-only cached scripts should be executed on slaves (requires Redis 7.0+)
Improvement - SORT_RO command is used for slave nodes (requires Redis 7.0+)
Improvement - decrease size of allocated data by RPermitExpirableSemaphore

Fixed - RedissonLocalCachedMap.clearLocalCache() method throws IllegalArgumentException
Fixed - RedissonMultiLock doesn't work properly with RedissonSpinLock
Fixed - SlaveConnectionPool no available Redis entries error occurs in Cluster mode
Fixed - RKeys.deleteByPattern() method does not always delete keys correctly
Fixed - expireAt(Instant) method of RExpirableReactive and RExpirableRx interfaces doesn't work
Fixed - wrong detection of added and removed slots in Redis Cluster mode
Fixed - RScoredSortedSet.addIfGreater() and RScoredSortedSet.addIfLess() methods always return false
Fixed - Spring Data Connection in multi mode causes thread stuck (regression since 3.16.7)
Fixed - Sentinel username setting is not applied (thanks to @nicolas-tg-ch)
Fixed - RTimeSeries doesn't handle same values for different timestamps
Fixed - Quarkus environment variables aren't parsed correctly
Fixed - check expiration before release in RPermitExpirableSemaphore (thanks to @randomVariable2)
Fixed - RedisTimeoutException: Command execution timeout for command: (PING) (regression since 3.16.3)
Fixed - wrong wait time calculation in RedissonMultiLock lock method causes deadlock
Fixed - RLocalCachedMap throws NPE if cache update listener receives message during init
Fixed - AsyncRemoteProxy throws Redisson is shutdown exception
Fixed - RedisClusterNode.clusterSlots() method throws Exception

redisson-3.16.8

Fixed - Quarkus redisson config fails to load in cluster mode with one node address
Fixed - registered RReliableTopic listener doesn't get old messages
Fixed - pubsub channel isn't released if subscription timeout occurred
Fixed - Quarkus Redisson config should be read at runtime
Fixed - RTopic channels aren't unsubscribed
Fixed - race condition causes Subscription timeout
Fixed - RMapCache.readAllKeySet() doesn't use MapKey codec
Fixed - Spring Data Redis RedissonConnection doesn't implement lpos command (thanks @woodyDM)
Fixed - master host isn't formatted into compressed format for IPV6 addresses in Sentinel mode
Fixed - Spring Data Redis restore() method throws Busy exception

redisson-3.16.7

Improvement - MessageListener should be annotated by FunctionalInterface

Fixed - RScript.scriptLoad() method doesn't load script into Slave nodes
Fixed - Spring Data RedissonConnection eval should use ByteArrayCodec (thanks @woodyDM)
Fixed - RSet.distributedIterator() and RScoredSortedSet.distributedIterator() methods throw script error
Fixed - synced slaves amount is not checked in RLock object
Fixed - race condition during hostname resolution in sentinel mode which may cause slave shutdown
Fixed - error should be thrown if slaves aren't defined in MasterSlave mode and readMode != MASTER
Fixed - master node shouldn't be initialized as slave in single mode
Fixed - can't find node error arise in replicated mode

redisson-3.16.6

Fixed - race condition causes wrong detection of failed slaves in Replicated mode. (regression since 3.16.5)

redisson-3.16.5

Feature - countIntersection() method added to RSet object
Feature - added reactive interface for RListMultimapCache and RSetMultimapCache objects
Feature - sentinelUsername setting added
Feature - added distributed iterator (thanks @Vorotyntsev)
Feature - added Spring Data Redis 2.6.0 support

Fixed - RedissonConnectionFactory.getReactiveConnection() method of Spring Data Redis isn't compatible with Redis cluster mode
Fixed - Mybatis RedissonCache should search redisson.yaml config at root package
Fixed - Can't find host in slaves! error after failover with a new IP of master host
Fixed - failed slaves aren't detected in Replicated mode
Fixed - get operation before put may cause inconsistent state of local cache
Fixed - RList.remove(object, count) throws exception if multiple objects were removed (thanks @cartermc24)
Fixed - RLocalCachedMap.delete() method clears local cache asynchronously
Fixed - IdleConnectionWatcher shouldn't close RedisPubSubConnection if it's in subscribed state
Fixed - SSL is not used for Sentinel master host
Fixed - update sync strategy of LocalCachedMap objects shouldn't apply updated value twice to instance of update source
Fixed - JCache dependency updated to 1.1.1
Fixed - Sentinel master-host = ? setting isn't handled properly during slave check
Fixed - RBuckets.trySet() method throws CROSSSLOT error (thanks to @deerRule)
Fixed - DNS monitor makes a new attempt to change master while current attempt wasn't finished

redisson-3.16.4

Feature - sentinelsDiscovery setting added
Feature - quarkus.redisson.file setting added to redisson-quarkus module to define external Redisson config file

Improvement - optimization of ClusterConnectionManager.checkSlaveNodesChange() and ClusterConnectionManager.checkMasterNodesChange() methods

Fixed - master change monitoring task in Replicated mode stops execution if it's invoked before the dns change
Fixed - RemoteService cannot be called if requestId is null (thanks to @jimichan)
Fixed - codec is not applied to RBuckets.set() method in non Cluster mode
Fixed - recovered slave shouldn't be added again in Redis Cluster mode
Fixed - releaseConnection method may cause StackOverflowError
Fixed - MOVED response with hostname isn't handled properly
Fixed - RStream.readGroup() method throws IndexOutOfBoundsException if group has a message without data
Fixed - NPE in CommandPubSubDecoder
Fixed - RExecutorService may execute same task twice at the same time
Fixed - dependencies for testing should use appropriate scope
Fixed - RPriorityQueue.add() method uses async method
Fixed - don't retry non-idempotent operations which were successfully sent
Fixed - RMapCache.fastRemove throws RedisException: too many results to unpack
Fixed - RRateLimiter decreases limit over the time in highly concurrent environment
Fixed - don't PING connection if it's in use

redisson-3.16.3

Improvement - RBuckets.get() method should group keys by slot in Redis Cluster mode
Improvement - RBatch result decoding optimization

Fixed - RExecutorService, RRemoteService execution may hang if connection used for tasks pooling was interrupted
Fixed - RBatch with skipResult() option affects result of other commands (regression since 3.16.1)
Fixed - connection leak (regression since 3.16.1)
Fixed - getBuckets().set() method throws CROSSSLOT error (thanks to @mikawudi)
Fixed - RedissonMapCache.addListener() method throws NPE
Fixed - master-host of Slave node isn't resolved in Sentinel mode
Fixed - interrupted RLock.tryLock() method keeps renewing lock indefinitely (thanks to @Cesarla)
Fixed - don't ping connection if it's in use
Fixed - natMapper isn't applied to resolved Sentinel and Cluster hosts

redisson-3.16.2

Feature - Micronaut 3.0 integration
Feature - added batched merge() method to RLiveObjectService interface
Feature - resolve hostnames used in Redis Cluster topology
Feature - resolve hostnames used in Redis Sentinel topology
Feature - added batched addLast() and addFirst() methods to RDeque, RDequeRx and RDequeReactive interfaces
Feature - added addAllCounted() and removeAllCounted() methods to RSet, RSetRx and RSetReactive interfaces

Fixed - Redis Stream trim command with MINID strategy is not fully supported
Fixed - Quarkus requires AutowiredAnnotationBeanPostProcessor class during native image execution
Fixed - issues with Quarkus Netty dependencies
Fixed - MOVED redirection loop detected error in Redis Cluster
Fixed - handling master with empty slots in Redis Cluster topology
Fixed - SentinelConnectionManager should use unified compressed format for IPv6
Fixed - RLocalCachedMap.readAllValues() method uses key decoder instead of value
Fixed - empty array passed to RKeys.delete() method causes thread blocking
Fixed - cluster partition without address causes NPE
Fixed - threads waiting for RSemaphore permits acquisition unable to acquire them if permits added
Fixed - RRateLimiter allows limit overcome
Fixed - RMapCacheReactive and RMapCacheRx interfaces miss method to define eviction algorithm
Fixed - write-behind tasks aren't flushed after Redisson shutdown() method invocation
Fixed - LiveObjects with indexed field can't be stored using batch persist method
Fixed - failed master shouldn't skipped in Redis Cluster topology scan (thanks to @JerryWzc)
Fixed - RListReactive iterator with filter returns non-deterministic result
Fixed - replicatedServers mode should use ip addresses if nodes defined using hostnames
Fixed - multiple masters check removed for replicatedServers mode
Fixed - MapWriter should be defined along with writeBehind settings