Keywords: Redis memory configuration | maxmemory parameter | OOM exception handling
Abstract: This article provides an in-depth exploration of Redis memory limit configuration, covering methods such as setting the maxmemory parameter via configuration files, dynamically adjusting memory limits using the CONFIG SET command, and persisting changes with CONFIG REWRITE. It explains the risks of the default setting (0 for unlimited memory) and offers examples of unit conversions from MB to GB. Additionally, the article addresses common OOM exceptions by emphasizing the importance of memory policies like allkeys-lru, and includes code examples to demonstrate how to prevent memory overflows in practical applications. Finally, best practices for configuration verification are summarized to ensure stable Redis operation under limited memory.
Core Mechanisms of Redis Memory Configuration
Memory management is a critical aspect of Redis performance optimization, with the maxmemory parameter allowing users to set the maximum memory usage for an instance. By default, this value is 0, indicating no memory limit, but this can lead to the operating system terminating the Redis process due to memory exhaustion; thus, it is strongly recommended to set a reasonable limit in production deployments. For example, adding maxmemory 2gb to the configuration file limits memory to 2GB. Redis supports various units, including bytes, KB, MB, and GB, such as maxmemory 2147483648 (byte representation of 2GB) or maxmemory 2gb (using extended units).
Dynamic Configuration and Persistence Methods
Beyond modifying configuration files, Redis offers dynamic configuration capabilities. The CONFIG SET maxmemory <value> command allows adjusting memory limits without restarting the service, e.g., CONFIG SET maxmemory 4294967296 sets a 4GB limit. To ensure changes persist after a restart, execute the CONFIG REWRITE command, which writes the current settings back to the configuration file. This approach is particularly useful for temporary adjustments or automation scripts in production environments.
Memory Policies and OOM Exception Handling
When memory usage reaches the maxmemory limit, Redis enforces an eviction policy based on the maxmemory-policy setting, such as allkeys-lru (Least Recently Used). If no policy is configured or memory is insufficient, OOM exceptions may occur, as seen in Java programs with errors like JedisDataException: OOM command not allowed when used memory > 'maxmemory'. To mitigate this, it is advisable to combine memory limits with policies, e.g., adding maxmemory-policy allkeys-lru to the configuration, and monitor memory usage through code:
import redis.clients.jedis.Jedis;
public class RedisMemoryMonitor {
public static void main(String[] args) {
Jedis jedis = new Jedis("localhost", 6379);
// Set memory limit to 40GB
jedis.configSet("maxmemory", "42949672960");
// Verify the setting
String maxMemory = jedis.configGet("maxmemory").get(1);
System.out.println("Max memory set to: " + maxMemory + " bytes");
// Check memory usage
String usedMemory = jedis.info("memory").split("\\n")[1].split(":")[1].trim();
System.out.println("Used memory: " + usedMemory + " bytes");
jedis.close();
}
}This code example demonstrates how to dynamically set and verify memory configurations using the Jedis client, while monitoring actual usage to help prevent OOM exceptions.
Configuration Verification and Best Practices
To ensure memory settings are effective, use the config get maxmemory command for verification, e.g., executing in Redis CLI returns 1) "maxmemory" 2) "42949672960" for a successful 40GB setup. Best practices include: always setting maxmemory to avoid system-level memory issues, combining it with maxmemory-policy for optimized eviction, and regularly monitoring memory trends with the INFO memory command. By following these steps, Redis stability and performance can be significantly enhanced.