Skip to content

Latest commit

 

History

History
232 lines (190 loc) · 10.1 KB

RELEASE-NOTES.md

File metadata and controls

232 lines (190 loc) · 10.1 KB
Raw

lettuce 4.4.0 RELEASE NOTES

This release ships with 76 tickets fixed.

It contains a range of performance improvements to lower GC pressure, introduces kqueue native transport on BSD-like systems and asynchronous, connection initialization for Redis Cluster and Cluster Pub/Sub subscriptions for a selection of nodes.

This release asserts compatibility in class-path mode with Java 9 requiring netty 4.1.11.Final.

Redis Cluster Pub/Sub on Node-Selections

RedisClusterClient.connectPubSub() now returns a StatefulRedisClusterPubSubConnection that allows registration of RedisClusterPubSubListeners and subscription on particular cluster nodes.

Cluster node-specific subscriptions allow usage of keyspace notifications. Keyspace notifications are different from userspace Pub/Sub since keyspace notifications don't broadcast to the entire cluster but are published only on the node the notification happens. A common usecase is when a key expires in the cluster.

StatefulRedisClusterPubSubConnection connection = client.connectPubSub();

connection.addListener(…);

connection.setNodeMessagePropagation(true);

RedisClusterPubSubCommands<String, String> sync = connection.sync();
sync.slaves().commands().psubscribe("__key*__:expire");

Native Transports

Lettuce now uses native transports by default, if the operating system is qualified and dependencies are available. Lettuce supports epoll (on Linux-based systems) since 4.0 and since this version kqueue (BSD-based systems, like macOS).

Epoll usage can be disabled with system properties by setting io.lettuce.core.epoll=false. In a similar way, kqueue can be disabled with io.lettuce.core.kqueue=false.

Epoll dependency:

<dependency>
    <groupId>io.netty</groupId>
    <artifactId>netty-transport-native-epoll</artifactId>
    <version>${netty-version}</version>
    <classifier>linux-x86_64</classifier>
</dependency>

Kqueue dependency:

<dependency>
    <groupId>io.netty</groupId>
    <artifactId>netty-transport-native-kqueue</artifactId>
    <version>${netty-version}</version>
    <classifier>osx-x86_64</classifier>
</dependency>

Asynchronous Connections in Redis Cluster

RedisClusterClient now connects asynchronously without intermediate synchronization to cluster nodes. The connection progress is shared between multiple threads that request a cluster node connection for the first time. Previously, connection was sequential-synchronous. Each connection attempt blocked subsequent attempts from other threads. If a cluster node connection ran into a timeout, threads were penalized with a growing wait time. If, say, 10 threads waited for a connection, the last thread had to wait up to 10 times the connection timeout.

Asynchronous connection initiates the connection once using internally a Future so multiple concurrent connection attempts share the resulting Future. Errors fail now faster and cluster node usage is fully asynchronous without synchronization and without the danger to run into a threading deadlock.

Scan Iterators

Scan operations (SCAN, HSCAN, SSCAN, ZSCAN) can be used through an enhanced ScanIterator API. Scan iterators implement Iterator allowing Java 8 stream usage and work with synchronous Redis Standalone and Redis Cluster connections. Scan iterators remove the complexity of cursor handling for initial/subsequent SCAN commands and do not require a different approach whether you use Redis Standalone or Redis Cluster.

RedisCommands<String, String> commands = …
ScanIterator<String> scan = ScanIterator.scan(commands, ScanArgs.Builder.limit(50).match("foo"));

// Iterator use
while(scan.hasNext()) {
    // …
}

// Java 8 Stream
scan.stream().filter(…).map(…).forEach(…)

Find the full change log at the end of this document.

Thanks to all contributors who made lettuce 4.4.0 possible. Lettuce 4.4.0 requires Java 8 or Java 9.

If you need any support, meet Lettuce at:

Enhancements

  • Allow Pub/Sub subscriptions on a node selection within Redis Cluster #385
  • Document RedisClient and RedisClusterClient implementation hooks #392 (Thanks to @HaloFour)
  • Provide scan iterators for SCAN/HSCAN/SSCAN and ZSCAN commands #397
  • Extend RedisClusterClient and RedisClusterFactoryBean to take multiple hosts #401
  • Default to epoll transport if available #402
  • Simplify NodeSelection implementation #408
  • Support CLIENT and INFO commands for Redis Sentinel #415
  • Allow setting client name with RedisURI #416
  • Extend RedisConnectionStateListener.onRedisConnected by adding SocketAddress #419
  • Connect asynchronously in cluster topology refresh #424 (Thanks to @szabowexler)
  • Propagate asynchronous connection initialization #429
  • Reuse stale connections collection when closing stale connections #443
  • Add ReadFrom.SLAVE_PREFERRED #452 (Thanks to @Spikhalskiy)
  • Avoid or reduce connection attempts to nodes in fail state #453 (Thanks to @Spikhalskiy)
  • Store ReadOnly commands and query these as bitmask #457
  • Initialize RedisStateMachine.LongProcessor in static initializer #481 (Thanks to @hellyguo)
  • Add shutdownAsync for RedisClient #485 (Thanks to @jongyeol)
  • Resolve hostnames using netty's DNS resolver #498
  • Introduce NettyCustomizer API #499
  • Use asynchronous connect in NodeSelection API #500
  • Gracefully degrade to successfully connected Cluster read candidates #503 (Thanks to @eliotw)
  • CommandHandler performs expensive queue check if requestQueueSize != Integer.MAX_VALUE #507 (Thanks to @CodingFabian)
  • Consider using ThreadLocalRandom instead of Random in RedisAdvancedCluster…CommandsImpl #510
  • Lots of memory occupied by firstInteger in CommandArgs #511 (Thanks to @CodingFabian)
  • Close/reconnect the connection when Redis protocol is corrupted #512 (Thanks to @jongyeol)
  • Allow setting the Redis password as char-array #513
  • Replace synchronized reconnect in ConnectionWatchdog with asynchronous connect #514
  • Align log prefix for CommandHandler and ConnectionWatchdog #538
  • Add support for kqueue transport #539
  • Add support for client certificate authentication #540
  • Reduce cross-thread queue sharing #544
  • Use pre-instantiated ByteBufProcessor in RedisStateMachine to find EOL #557

Fixes

  • Fail to second call pool.borrowObject() using StatefulRedisMasterSlaveConnection in 4.3.0.Final #411 (Thanks to @krisjey)
  • Deprecate addListener/removeListener methods from execution-specific Pub/Sub interfaces. #412
  • Unexpected 'Invalid database number' exception for database > 15 #420 (Thanks to @ChitraGuru)
  • Dynamic API: array of arguments are not supported #421 (Thanks to @coolmeen)
  • ClusterClientOptions.socketOptions returns ClientOptions.Builder #425 (Thanks to @maxd)
  • Lettuce suppresses initial exception in Redis Cluster #427 (Thanks to @maxd)
  • CDI can't create RedisClusterClient because of missing default constructor #438 (Thanks to @wuhuaxu)
  • RoundRobinSocketAddressSupplier may contain more RedisClusterNodes than the current topology view #440
  • Partitions.addPartition and reload not correctly synchronized #442
  • pingBeforeActivateConnection and authentication fails using Redis Sentinel #448 (Thanks to @coolmeen)
  • PooledClusterConnectionProvider.close does not close connections #460
  • Possible race condition in ReconnectHandler.prepareClose() #465
  • Dynamic API: Interface implementation fails for non-string connections #467
  • ChannelInitializer (PING before connect) waits indefinite #470
  • RedisClient.shutdown() does not synchronize on clientResources.shutdown future #475
  • Unknown channel option warnings in log using UDS #476 (Thanks to @magdkudama)
  • BlockingOperationException using RedisClusterClient with rxjava keys operation #486 (Thanks to @jeniok)
  • Latency metrics is reset even if set to false #487 (Thanks to @ameenhere)
  • firstResponseLatency depends on command completion time #491
  • Shutdown PauseDetector if DefaultCommandLatencyCollector is shut down and PauseDetector is no longer required #516
  • LatencyStats leaves PauseDetectorListener registered #517
  • Shaded JAR contains too many dependencies #519
  • Fix static initialization of PauseDetector when LatencyUtils are not on the class path #520
  • jcl-over-slf4j is a compile dependency instead of test dependency #526
  • NumberFormatException on "inf" zscore #528 (Thanks to @DarkSeraphim)
  • Connection pool deadlock #531 (Thanks to @DarkSeraphim)
  • Improve Epoll availability detection #535
  • Close connections in pooling destroy hook #545
  • Adopt to changed SLOWLOG output #551
  • CommandHandler.write contains isConnected verification #556

Other

  • Test with JDK 9 Early Access builds #310
  • Fix README #410 (Thanks to @changeworld)
  • Annotate deprecated items with @Deprecated #414
  • Disable WAIT as Read-Only command #418
  • Upgrade to netty 4.1.7/4.0.43 #436
  • Fix ClientResources usage in tests #445
  • Enable JMH tests on TravisCI #454
  • Upgrade to TravisCI trusty #461
  • Upgrade to Mockito 2 #469
  • Improve JavaDoc #472
  • Upgrade dependencies #474
  • Switch optional dependencies from provided to optional #482
  • Update Netty version to 4.1.9 #489 (Thanks to @odiszapc)
  • Move to lettuce.io #493
  • Switch to codecov #504
  • Upgrade integrations from mp911de/lettuce to lettuce-io/lettuce-core #506
  • Upgrade to RxJava 1.2.10 #524
  • Upgrade to netty 4.1.10 #532
  • Upgrade to netty 4.0.11 #537
  • Upgrade to netty 4.1.12 #548
  • Allow Redis version pinning for build #552

Lettuce requires a minimum of Java 8 to build and run. It is tested continuously against the latest Redis source-build.

If you need any support, meet Lettuce at