Connecting to Redis on Redisson (Java)
This section describes how to access a Redis instance on Redisson. For more information about how to use other Redis clients, visit the Redis official website.
For Spring Boot projects, Spring Data Redis is already integrated with Jedis and Lettuce, but does not support Redisson. Redisson provides the redisson-spring-boot-starter component (https://mvnrepository.com/artifact/org.redisson/redisson) that can be used with Spring Boot.
Spring Boot 1.x is integrated with Jedis, and Spring Boot 2.x is integrated with Lettuce.
- If a password was set during DCS Redis instance creation, configure the password for connecting to Redis using Redisson. Do not hard code the plaintext password.
- To connect to a single-node, read/write splitting, or Proxy Cluster instance, use the useSingleServer method of the SingleServerConfig object of Redisson. To connect to a master/standby instance, use the useMasterSlaveServers method of the MasterSlaveServersConfig object of Redisson. To connect to a Redis Cluster instance, use the useClusterServers method of the ClusterServersConfig object.
- Springboot 2.3.12.RELEASE or later is required. Redisson 3.37.0 or later is required.
Prerequisites
- A Redis instance is created, and is in the Running state. To create a Redis instance, see Buying a DCS Redis Instance.
- View the IP address/domain name and port number of the DCS Redis instance to be accessed. For details, see Viewing and Modifying Basic Settings of a DCS Instance.
Pom Configuration
<!-- spring-data-redis --> <dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-data-redis</artifactId> <exclusions> <!--Lettuce is integrated in Spring Boot 2.x by default. This dependency needs to be deleted. --> <exclusion> <artifactId>lettuce-core</artifactId> <groupId>io.lettuce</groupId> </exclusion> </exclusions> </dependency> <!--Redisson's adaptation package for Spring Boot--> <dependency> <groupId>org.redisson</groupId> <artifactId>redisson-spring-boot-starter</artifactId> <version>${redisson.version}</version> </dependency>
Bean Configuration
Spring Boot does not provide Redisson adaptation, and the application.properties configuration file does not have the corresponding configuration item. Therefore, you can only use Bean configuration.
- Single-node, read/write splitting, and Proxy Cluster
import org.redisson.Redisson; import org.redisson.api.RedissonClient; import org.redisson.codec.JsonJacksonCodec; import org.redisson.config.Config; import org.redisson.config.SingleServerConfig; import org.springframework.beans.factory.annotation.Value; import org.springframework.context.annotation.Bean; import org.springframework.context.annotation.Configuration; @Configuration public class SingleConfig { @Value("${redis.address:}") private String redisAddress; @Value("${redis.password:}") private String redisPassword; @Value("${redis.database:0}") private Integer redisDatabase = 0; @Value("${redis.connect.timeout:3000}") private Integer redisConnectTimeout = 3000; @Value("${redis.connection.idle.timeout:10000}") private Integer redisConnectionIdleTimeout = 10000; @Value("${redis.connection.ping.interval:1000}") private Integer redisConnectionPingInterval = 1000; @Value("${redis.timeout:2000}") private Integer timeout = 2000; @Value("${redis.connection.pool.min.size:50}") private Integer redisConnectionPoolMinSize; @Value("${redis.connection.pool.max.size:200}") private Integer redisConnectionPoolMaxSize; @Value("${redis.retry.attempts:3}") private Integer redisRetryAttempts = 3; @Value("${redis.retry.interval:200}") private Integer redisRetryInterval = 200; @Bean public RedissonClient redissonClient(){ Config redissonConfig = new Config(); SingleServerConfig serverConfig = redissonConfig.useSingleServer(); serverConfig.setAddress(redisAddress); serverConfig.setConnectionMinimumIdleSize(redisConnectionPoolMinSize); serverConfig.setConnectionPoolSize(redisConnectionPoolMaxSize); serverConfig.setDatabase(redisDatabase); serverConfig.setPassword(redisPassword); serverConfig.setConnectTimeout(redisConnectTimeout); serverConfig.setIdleConnectionTimeout(redisConnectionIdleTimeout); serverConfig.setPingConnectionInterval(redisConnectionPingInterval); serverConfig.setTimeout(timeout); serverConfig.setRetryAttempts(redisRetryAttempts); serverConfig.setRetryInterval(redisRetryInterval); redissonConfig.setCodec(new JsonJacksonCodec()); return Redisson.create(redissonConfig); } }
- Master/Standby
import org.redisson.Redisson; import org.redisson.api.RedissonClient; import org.redisson.codec.JsonJacksonCodec; import org.redisson.config.Config; import org.redisson.config.MasterSlaveServersConfig; import org.redisson.config.ReadMode; import org.redisson.config.SubscriptionMode; import org.springframework.beans.factory.annotation.Value; import org.springframework.context.annotation.Bean; import org.springframework.context.annotation.Configuration; import java.util.HashSet; @Configuration public class MasterStandbyConfig { @Value("${redis.master.address}") private String redisMasterAddress; @Value("${redis.slave.address}") private String redisSlaveAddress; @Value("${redis.database:0}") private Integer redisDatabase = 0; @Value("${redis.password:}") private String redisPassword; @Value("${redis.connect.timeout:3000}") private Integer redisConnectTimeout = 3000; @Value("${redis.connection.idle.timeout:10000}") private Integer redisConnectionIdleTimeout = 10000; @Value("${redis.connection.ping.interval:1000}") private Integer redisConnectionPingInterval = 1000; @Value("${redis.timeout:2000}") private Integer timeout = 2000; @Value("${redis.master.connection.pool.min.size:50}") private Integer redisMasterConnectionPoolMinSize = 50; @Value("${redis.master.connection.pool.max.size:200}") private Integer redisMasterConnectionPoolMaxSize = 200; @Value("${redis.retry.attempts:3}") private Integer redisRetryAttempts = 3; @Value("${redis.retry.interval:200}") private Integer redisRetryInterval = 200; @Bean public RedissonClient redissonClient() { Config redissonConfig = new Config(); MasterSlaveServersConfig serverConfig = redissonConfig.useMasterSlaveServers(); serverConfig.setMasterAddress(redisMasterAddress); HashSet<String> slaveSet = new HashSet<>(); slaveSet.add(redisSlaveAddress); serverConfig.setSlaveAddresses(slaveSet); serverConfig.setDatabase(redisDatabase); serverConfig.setPassword(redisPassword); serverConfig.setMasterConnectionMinimumIdleSize(redisMasterConnectionPoolMinSize); serverConfig.setMasterConnectionPoolSize(redisMasterConnectionPoolMaxSize); serverConfig.setReadMode(ReadMode.MASTER); serverConfig.setSubscriptionMode(SubscriptionMode.MASTER); serverConfig.setConnectTimeout(redisConnectTimeout); serverConfig.setIdleConnectionTimeout(redisConnectionIdleTimeout); serverConfig.setPingConnectionInterval(redisConnectionPingInterval); serverConfig.setTimeout(timeout); serverConfig.setRetryAttempts(redisRetryAttempts); serverConfig.setRetryInterval(redisRetryInterval); redissonConfig.setCodec(new JsonJacksonCodec()); return Redisson.create(redissonConfig); } }
- Redis Cluster
import org.redisson.Redisson; import org.redisson.api.RedissonClient; import org.redisson.codec.JsonJacksonCodec; import org.redisson.config.ClusterServersConfig; import org.redisson.config.Config; import org.redisson.config.ReadMode; import org.redisson.config.SubscriptionMode; import org.springframework.beans.factory.annotation.Value; import org.springframework.context.annotation.Bean; import org.springframework.context.annotation.Configuration; import java.util.List; @Configuration public class ClusterConfig { @Value("${redis.cluster.address}") private List<String> redisClusterAddress; @Value("${redis.cluster.scan.interval:5000}") private Integer redisClusterScanInterval = 5000; @Value("${redis.password:}") private String redisPassword; @Value("${redis.connect.timeout:3000}") private Integer redisConnectTimeout = 3000; @Value("${redis.connection.idle.timeout:10000}") private Integer redisConnectionIdleTimeout = 10000; @Value("${redis.connection.ping.interval:1000}") private Integer redisConnectionPingInterval = 1000; @Value("${redis.timeout:2000}") private Integer timeout = 2000; @Value("${redis.retry.attempts:3}") private Integer redisRetryAttempts = 3; @Value("${redis.retry.interval:200}") private Integer redisRetryInterval = 200; @Value("${redis.master.connection.pool.min.size:50}") private Integer redisMasterConnectionPoolMinSize = 50; @Value("${redis.master.connection.pool.max.size:200}") private Integer redisMasterConnectionPoolMaxSize = 200; @Bean public RedissonClient redissonClient() { Config redissonConfig = new Config(); ClusterServersConfig serverConfig = redissonConfig.useClusterServers(); serverConfig.setNodeAddresses(redisClusterAddress); serverConfig.setScanInterval(redisClusterScanInterval); serverConfig.setPassword(redisPassword); serverConfig.setMasterConnectionMinimumIdleSize(redisMasterConnectionPoolMinSize); serverConfig.setMasterConnectionPoolSize(redisMasterConnectionPoolMaxSize); serverConfig.setReadMode(ReadMode.MASTER); serverConfig.setSubscriptionMode(SubscriptionMode.MASTER); serverConfig.setConnectTimeout(redisConnectTimeout); serverConfig.setIdleConnectionTimeout(redisConnectionIdleTimeout); serverConfig.setPingConnectionInterval(redisConnectionPingInterval); serverConfig.setTimeout(timeout); serverConfig.setRetryAttempts(redisRetryAttempts); serverConfig.setRetryInterval(redisRetryInterval); redissonConfig.setCodec(new JsonJacksonCodec()); return Redisson.create(redissonConfig); } }
Parameter Description
Parameter |
Default Value |
Description |
---|---|---|
codec |
org.redisson.codec.JsonJacksonCodec |
Encoding format, including JSON, Avro, Smile, CBOR, and MsgPack. |
threads |
Number of CPU cores x 2 |
Thread pool used for executing RTopic Listener, RRemoteService, and RExecutorService. |
executor |
null |
The function is the same as threads. If this parameter is not set, a thread pool is initialized based on threads. |
nettyThreads |
Number of CPU cores x 2 |
Thread pool used by the TCP channel that connects to the redis-server. All channels share this connection pool and are mapped to Netty's Bootstrap.group(...). |
eventLoopGroup |
null |
The function is the same as nettyThreads. If this parameter is not set, an EventLoopGroup is initialized based on the nettyThreads parameter for the bottom-layer TCP channel to use. |
transportMode |
TransportMode.NIO |
Transmission mode. The options are NIO, EPOLL (additional package required), and KQUEUE (additional package required). |
lockWatchdogTimeout |
30000 |
Timeout interval (in milliseconds) of the lock-monitoring watchdog. In the distributed lock scenario, if the leaseTimeout parameter is not specified, the default value of this parameter is used. |
keepPubSubOrder |
true |
Indicates whether to receive messages in the publish sequence. If messages can be processed concurrently, you are advised to set this parameter to false. |
Parameter |
Default Value |
Description |
---|---|---|
address |
- |
Node connection information, in redis://ip:port format. |
database |
0 |
ID of the database to be used. |
connectionMinimumIdleSize |
32 |
Minimum number of connections to the master node of each shard. |
connectionPoolSize |
64 |
Maximum number of connections to the master node of each shard. |
subscriptionConnectionMinimumIdleSize |
1 |
Minimum number of connections to the target node for pub/sub. |
subscriptionConnectionPoolSize |
50 |
Maximum number of connections to the target node for pub/sub. |
subcriptionPerConnection |
5 |
Maximum number of subscriptions on each subscription connection. |
connectionTimeout |
10000 |
Connection timeout interval, in milliseconds. |
idleConnectionTimeout |
10000 |
Maximum time (in milliseconds) for reclaiming idle connections. |
pingConnectionInterval |
30000 |
Heartbeat for detecting available connections, in milliseconds. Recommended: 3000 ms. |
timeout |
3000 |
Timeout interval for waiting for a response, in milliseconds. |
retryAttempts |
3 |
Maximum number of retries upon send failures. |
retryInterval |
1500 |
Retry interval, in milliseconds. Recommended: 200 ms. |
clientName |
null |
Client name. |
Parameter |
Default Value |
Description |
---|---|---|
masterAddress |
- |
Master node connection information, in redis://ip:port format. |
slaveAddresses |
- |
Standby node connection information, in Set<redis://ip:port> format. |
readMode |
SLAVE |
Read mode. By default, read traffic is distributed to replica nodes. The value can be MASTER (recommended), SLAVE, or MASTER_SLAVE. Other values may cause access failures in failover scenarios. |
loadBalancer |
RoundRobinLoadBalancer |
Load balancing algorithm. This parameter is valid only when readMode is set to SLAVE or MASTER_SLAVE. Read traffic is distributed evenly. |
masterConnectionMinimumIdleSize |
32 |
Minimum number of connections to the master node of each shard. |
masterConnectionPoolSize |
64 |
Maximum number of connections to the master node of each shard. |
slaveConnectionMinimumIdleSize |
32 |
Minimum number of connections to each replica node of each shard. If readMode is set to MASTER, the value of this parameter is invalid. |
slaveConnectionPoolSize |
64 |
Maximum number of connections to each replica node of each shard. If readMode is set to MASTER, the value of this parameter is invalid. |
subscriptionMode |
SLAVE |
Subscription mode. By default, only replica nodes handle subscription. The value can be SLAVE or MASTER (recommended). |
subscriptionConnectionMinimumIdleSize |
1 |
Minimum number of connections to the target node for pub/sub. |
subscriptionConnectionPoolSize |
50 |
Maximum number of connections to the target node for pub/sub. |
subcriptionPerConnection |
5 |
Maximum number of subscriptions on each subscription connection. |
connectionTimeout |
10000 |
Connection timeout interval, in milliseconds. |
idleConnectionTimeout |
10000 |
Maximum time (in milliseconds) for reclaiming idle connections. |
pingConnectionInterval |
30000 |
Heartbeat for detecting available connections, in milliseconds. Recommended: 3000 ms. |
timeout |
3000 |
Timeout interval for waiting for a response, in milliseconds. |
retryAttempts |
3 |
Maximum number of retries upon send failures. |
retryInterval |
1500 |
Retry interval, in milliseconds. Recommended: 200 ms. |
clientName |
null |
Client name. |
Parameter |
Default Value |
Description |
---|---|---|
nodeAddress |
- |
Connection addresses of cluster nodes. Each address uses the redis://ip:port format. Use commas (,) to separate connection addresses of different nodes. |
password |
null |
Password for logging in to the cluster. |
scanInterval |
1000 |
Interval for periodically checking the cluster node status, in milliseconds. |
readMode |
SLAVE |
Read mode. By default, read traffic is distributed to replica nodes. The value can be MASTER (recommended), SLAVE, or MASTER_SLAVE. Other values may cause access failures in failover scenarios. |
loadBalancer |
RoundRobinLoadBalancer |
Load balancing algorithm. This parameter is valid only when readMode is set to SLAVE or MASTER_SLAVE. Read traffic is distributed evenly. |
masterConnectionMinimumIdleSize |
32 |
Minimum number of connections to the master node of each shard. |
masterConnectionPoolSize |
64 |
Maximum number of connections to the master node of each shard. |
slaveConnectionMinimumIdleSize |
32 |
Minimum number of connections to each replica node of each shard. If readMode is set to MASTER, the value of this parameter is invalid. |
slaveConnectionPoolSize |
64 |
Maximum number of connections to each replica node of each shard. If readMode is set to MASTER, the value of this parameter is invalid. |
subscriptionMode |
SLAVE |
Subscription mode. By default, only replica nodes handle subscription. The value can be SLAVE or MASTER (recommended). |
subscriptionConnectionMinimumIdleSize |
1 |
Minimum number of connections to the target node for pub/sub. |
subscriptionConnectionPoolSize |
50 |
Maximum number of connections to the target node for pub/sub. |
subcriptionPerConnection |
5 |
Maximum number of subscriptions on each subscription connection. |
connectionTimeout |
10000 |
Connection timeout interval, in milliseconds. |
idleConnectionTimeout |
10000 |
Maximum time (in milliseconds) for reclaiming idle connections. |
pingConnectionInterval |
30000 |
Heartbeat for detecting available connections, in milliseconds. Recommended: 3000. |
timeout |
3000 |
Timeout interval for waiting for a response, in milliseconds. |
retryAttempts |
3 |
Maximum number of retries upon send failures. |
retryInterval |
1500 |
Retry interval, in milliseconds. Recommended: 200. |
clientName |
null |
Client name. |
Suggestion for Configuring DCS Instances
- readMode
MASTER is the recommended value, that is, the master node bears all read and write traffic. This is to avoid data inconsistency caused by master/replica synchronization latency. If the value is SLAVE, all read requests will trigger errors when replicas are faulty. If the value is MASTER_SLAVE, some read requests will trigger errors. Read errors last for the period specified by failedSlaveCheckInterval (180s by default) until the faulty nodes are removed from the available node list.
If read traffic and write traffic need to be separated, you can use read/write splitting DCS instances. Proxy nodes are deployed in the middle to distribute read and write traffic. When a replica node is faulty, traffic is automatically switched to the master node. The switchover does not interrupt service applications, and the fault detection time window is far shorter than Redisson's window.
- subscriptionMode
Similar to readMode, MASTER is the recommended value.
- Connection pool configuration
The following calculation is applicable only to common service scenarios. You can customize it based on your service requirements.
There is no standard connection pool size. You can configure one based on your service traffic. The following formulas are for reference:
- Minimum number of connections = (QPS of a single node accessing Redis)/(1000 ms/Average time spent on a single command)
- Maximum number of connections = (QPS of a single node accessing Redis)/(1000 ms/Average time spent on a single command) x 150%
For example, if the QPS of a service application is about 10,000, each request needs to access Redis 10 times (that is, 100,000 accesses to Redis every second), and the service application has 10 hosts, the calculation is as follows:
QPS of a single node accessing Redis = 100,000/10 = 10,000
Average time spent on a single command = 20 ms (Redis takes 5 ms to 10 ms to process a single command under normal conditions. If network jitter occurs, it takes 15 ms to 20 ms.)
Minimum number of connections = 10,000/(1000 ms/20 ms) = 200
Maximum number of connections = 10,000/(1000 ms/20 ms) × 150% = 300
- Retry configuration
Redisson supports retries. You can set the following parameters based on service requirements. Generally, configure three retries, and set the retry interval to about 200 ms.
- retryAttempts: number of retry times
- retryInterval: retry interval
In Redisson, some APIs are implemented through LUA, and the performance is low. You are advised to use Jedis instead of Redisson.
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.