触发器和函数示例

如何使用触发器和函数

Redis 堆栈 Redis 社区版 Redis 企业软件 Redis 云 Redis 社区版 适用于 Kubernetes 的 Redis Enterprise 客户

Redis Stack 的触发器和函数功能可以在数据发生更改时检测到这些更改，并保证在数据源中无缝执行业务逻辑。这确保了作数据的新选项是可能的，并且可以同时为所有客户端提供，同时简化了部署和维护。让我们探索一些特定于行业的使用案例，这些功能在这些用例中大放异彩：

  • 零售：在零售领域，可以开发一个功能，在收到订单后立即更新库存。这使企业能够准确预测任何给定日期特定仓库的库存需求。
  • 旅游：对于旅游行业，可以使用触发器来检测新的航班预订，并有效地将相关信息加载到不同消费者的队列中。然后，服务可以利用这些数据为酒店、餐馆、汽车租赁服务等提供建议。利用 Redis 几何图形，强大的推荐系统可以提供本地化和个性化的建议。
  • 订阅服务：在订阅服务领域，使用密钥空间触发器可以自动识别订阅已续订的用户，从而无缝将其状态更改为活动状态。可以对这些用户执行进一步的作，例如将他们添加到队列中以发送通知或执行其他作。

这些示例突出了触发器和功能在不同行业中的实际应用，展示了它们在简化流程和提供高效解决方案方面的价值。

使用键空间和 Stream 触发器来捕获事件并执行所需的 JavaScript 函数需要几行代码来丰富应用程序，从而使用新行为来丰富应用程序。以下示例显示了如何使用 JavaScript 解决典型问题。

审计和日志记录

Redis 触发器可用于检测对特定数据结构的更改，并在 Redis 日志中记录审计跟踪。最新更改的时间戳可以在同一函数中更新。

function alertUserChanged(client, data) {
    // detect the event and log an audit trail
    if (data.event == 'hset'){
        redis.log('User data changed: ' + data.key);
    }
    else if (data.event == 'del'){
        redis.log('User deleted: ' + data.key);
    }

    var curr_time = client.call("time")[0];

    // add the current timestamp to the Hash
    client.call('hset', data.key, 'last', curr_time);
}

redis.registerKeySpaceTrigger('alert_user_changed', 'user:', 
    alertUserChanged, {description: 'Report user data change.'});

Enrich and transform data

Data can be extracted, enriched or transformed, and loaded again. As an example, upon insertion of a document in a Hash data structure, a Trigger can launch the execution of a Function that computes the number of words in the text (in the example, a simple tokenization is presented but the logic can be as complex as required). The counter is finally stored in the same Hash together with the original document.

function wordsCounter(client, data){
    text = client.call('hget', data.key, 'content');
    words = text.split(' ').length;
    client.call('hset', data.key, 'cnt', words.toString());
    redis.log('Number of words: ' + words.toString()); //This log is for demo purposes, be aware of spamming the log file in production
}

redis.registerKeySpaceTrigger('words_counter', 'doc:', 
    wordsCounter, {description: 'Count words in a document.'});

Batch operations

JavaScript functions can be executed when required, for example as part of scheduled or periodic maintenance routines. An example could be deleting data identified by the desired pattern.

#!js api_version=1.0 name=utils


redis.registerAsyncFunction('del_keys', async function(async_client, pattern){
    var count = 0;
    var cursor = '0';
    do {
        async_client.block((client)=>{
            var res = client.call('scan', cursor, 'match', pattern);
            cursor = res[0];
            var keys = res[1];
            keys.forEach((key) => {
                client.call('del', key);
            });
        });
    } while(cursor != '0');
    return count;
});

The function del_keys performs an asynchronous batch scan of all the keys in the database matching the desired pattern, and enters a blocking section where it gets partial results. The asynchronous design of this function permits non-blocking execution, so other clients will not starve while the iteration is performed on a keyspace of arbitrary size. The function can be invoked as follows.

TFCALLASYNC utils.del_keys 0 "user:*"

Automatic expiration

Sometimes it is useful to enforce expiration time for data that by nature is short-lived. An example could be session data, or authentication tokens. A trigger fits the use case and can execute a function that sets the desired TTL.

function automaticSessionExpiry(client, data){
    client.call('expire', data.key, '3600');
    redis.log('Session ' + data.key + ' has been automatically expired'); //This log is for demo purposes, be aware of spamming the log file in production
}

redis.registerKeySpaceTrigger('automaticSessionExpiry', 'session:', 
    automaticSessionExpiry, {description: 'Set the session TTL to one hour.'});
RATE THIS PAGE
Back to top ↑