跳到主要内容
版本:4.x

RocketMQ Prometheus Exporter

介绍

Rocketmq-exporter 是用于监控 RocketMQ broker 端和客户端所有相关指标的系统,通过 mqAdmin 从 broker 端获取指标值后封装成 87 个 cache。

警告

过去版本曾是 87 个 concurrentHashMap,由于 Map 不会删除过期指标,所以一旦有 label 变动就会生成一个新的指标,旧的无用指标无法自动删除,久而久之造成内存溢出。而使用 Cache 结构可可以实现过期删除,且过期时间可配置。

Rocketmq-expoter 获取监控指标的流程如下图所示,Expoter 通过 MQAdminExt 向 MQ 集群请求数据,请求到的数据通过 MetricService 规范化成 Prometheus 需要的格式,然后通过 /metics 接口暴露给 Prometheus。 165754739545

Metric 结构

Metric 类位于 org.apache.rocketmq.expoter.model.metrics 包下,实质上是一些实体类,每个实体类代表一类指标, 总共 14 个 Metric 类。这些类作为 87 个 Cache 的 key, 用不同的 label 值进行区分。

实体类中包含了 label 的三个维度:broker、consumer、producer
  • broker 相关 metric 类有: BrokerRuntimeMetric、BrokerMetric、DLQTopicOffsetMetric、TopicPutNumMetric

  • 消费者相关类有: ConsumerRuntimeConsumeFailedMsgsMetric 、ConsumerRuntimeConsumeFailedTPSMetric 、ConsumerRuntimeConsumeOKTPSMetric、ConsumerRuntimeConsumeRTMetric、ConsumerRuntimePullRTMetric、ConsumerRuntimePullTPSMetric、ConsumerCountMetric、ConsumerMetric、ConsumerTopicDiffMetric

  • 生产者相关 metric 类有: ProducerMetric

Prometheus 拉取 metrics 的过程

RocketMQ-exporter 项目和 Prometheus 相当于服务器和客户端的关系,RocketMQ-exporter 项目引入了 Prometheus 的 client 包,该包中规定了需要获取的信息的类型即项目中的 MetricFamilySamples 类,Prometheus 向 expoter 请求 metrics,expoter 将信息封装成相应的类型之后返回给 Prometheus。

rocketmq-expoter 项目启动后,会获取 rocketmq 的各项 metrics 收集到 mfs 对象中,当浏览器或 Prometheus 访问相应的接口时,会通过 service 将 mfs 对象中的 samples 生成 Prometheus 所支持的格式化数据。主要包含以下步骤:

浏览器通过访问 ip:5557/metrics,会调用 RMQMetricsController 类下的 metrics 方法,其中 ip 为 rocketmq-expoter 项目运行的主机 ip

private void metrics(HttpServletResponse response) throws IOException {
StringWriter writer = new StringWriter();
metricsService.metrics(writer);
response.setHeader("Content-Type", "text/plain; version=0.0.4; charset=utf-8");
response.getOutputStream().print(writer.toString());
}

通过新建 StringWriter 对象用于收集 metrics 指标,调用 MetricsService 类中的方法 metrics 将 expoter 中提取到的指标收集到 writer 对象中,最后将收集到的指标输出到网页上。

收集到的指标格式为:

<metric name>{<label name>=<label value>, ...} <metric value>

如:

rocketmq_group_diff{group="rmq_group_test_20220114",topic="fusion_console_tst",countOfOnlineConsumers="0",msgModel="1",} 23.0

MetricCollectTask 类中的 5 个定时任务

MetricCollectTask 类中有 5 个定时任务,分别为 collectTopicOffset、collectConsumerOffset、collectBrokerStatsTopic、collectBrokerStats 和 collectBrokerRuntimeStats。用于收集消费位点信息以及 Broker 状态信息等。其 cron 表达式为:cron: 15 0/1 * * * ?,表示每分钟会收集一次。其核心功能是通过 mqAdminExt 对象从集群中获取 broker 中的信息,然后将其添加到对应的 87 个监控指标中,以 collectTopicOffset 为例:

  1. 首先初始化TopicList对象,通过mqAdminExt.fetchAllTopicList()方法获取到集群的所有topic信息。
   ```java
TopicList topicList = null;
try { topicList = mqAdminExt.fetchAllTopicList();
} catch (Exception ex) {
log.error(String.format("collectTopicOffset-exception comes getting topic list from namesrv, address is %s",
JSON.toJSONString(mqAdminExt.getNameServerAddressList())));
return;
}
  1. 将 topic 加入到 topicSet 中,循环遍历每一个 topic,通过 mqAdminExt.examineTopicStats(topic)函数来检查 topic 状态。
    Set < String > topicSet = topicList != null ? topicList.getTopicList() : null;
for (String topic: topicSet) {
TopicStatsTable topicStats = null;
try {
topicStats = mqAdminExt.examineTopicStats(topic);
} catch (Exception ex) {
log.error(String.format("collectTopicOffset-getting topic(%s) stats error. the namesrv address is %s",
topic,
JSON.toJSONString(mqAdminExt.getNameServerAddressList())));
continue;}
  1. 初始化 topic 状态 set,用于用于按 broker 划分的 topic 信息位点的 hash 表 brokerOffsetMap,以及一个用于按 broker 名字为 key 的用于存储更新时间戳的 hash 表 brokerUpdateTimestampMap。
        Set<Map.Entry<MessageQueue, TopicOffset>> topicStatusEntries = topicStats.getOffsetTable().entrySet();
HashMap<String, Long> brokerOffsetMap = new HashMap<>();
HashMap<String, Long> brokerUpdateTimestampMap = new HashMap<>();
for (Map.Entry<MessageQueue, TopicOffset> topicStatusEntry : topicStatusEntries) {
MessageQueue q = topicStatusEntry.getKey();
TopicOffset offset = topicStatusEntry.getValue();
if (brokerOffsetMap.containsKey(q.getBrokerName())) {
brokerOffsetMap.put(q.getBrokerName(), brokerOffsetMap.get(q.getBrokerName()) + offset.getMaxOffset());
} else {
brokerOffsetMap.put(q.getBrokerName(), offset.getMaxOffset());
}
if (brokerUpdateTimestampMap.containsKey(q.getBrokerName())) {
if (offset.getLastUpdateTimestamp() > brokerUpdateTimestampMap.get(q.getBrokerName())) {
brokerUpdateTimestampMap.put(q.getBrokerName(), offset.getLastUpdateTimestamp());
}
} else {
brokerUpdateTimestampMap.put(q.getBrokerName(),
offset.getLastUpdateTimestamp());
}
}

  1. 最后通过遍历 brokerOffsetMap 中的每一项,通过调用 metricsService 获取到 metricCollector 对象,调用 RMQMetricsCollector 类中的 addTopicOffsetMetric 方法,将相应的值添加到 RMQMetricsCollector 类中 87 个指标对应的其中一个指标的 cache 中。
 Set<Map.Entry<String, Long>> brokerOffsetEntries = brokerOffsetMap.entrySet();
for (Map.Entry<String, Long> brokerOffsetEntry : brokerOffsetEntries) {
metricsService.getCollector().addTopicOffsetMetric(clusterName, brokerOffsetEntry.getKey(), topic,
brokerUpdateTimestampMap.get(brokerOffsetEntry.getKey()), brokerOffsetEntry.getValue());
}
}
log.info("topic offset collection task finished...." + (System.currentTimeMillis() - start));
}

Rocketmq-exporter 收集指标流程图

dafcvznvkldaf

快速开始

配置 application.yml

application.yml 中重要的配置主要有:

  • server.port 设置 prometheus 监听 rocketmq-exporter 的端口, 默认为 5557

  • rocketmq.config.webTelemetryPath 配置 prometheus 获取指标的路径,默认为 /metrics ,使用默认值即可.

  • rocketmq.config.enableACL 如果 RocketMQ 集群开启了 ACL 验证,需要配置为 true, 并在 accessKey 和 secretKey 中配置相应的 ak, sk.

  • rocketmq.config.outOfTimeSeconds 用于配置存储指标和相应的值的过期时间,若超过该时间,cache 中的 key 对应的节点没有发生写更改,则会进行删除.一般配置为 60s 即可(根据 prometheus 获取指标的时间间隔进行合理配置,只要保证过期时间大于等于 prometheus 收集指标的时间间隔即可)

  • task..cron 配置 exporter 从 broker 拉取指标的定时任务的时间间隔,默认值为"15 0/1 * * ?" 每分钟的 15s 拉取一次指标.

启动 exporter 项目

按照 prometheus 官网配置启动

配置 prometheus 的 static_config: -targets 为 exporter 的启动 IP 和端口,如: localhost:5557

访问 prometheus 页面

本地启动默认为: localhost:9090 ,则可对收集到的指标值进行查看,如下图所示:

3298559603

提示

为了达到更好的可视化效果,观察指标值变化趋势, prometheus 搭配 grafana 效果更佳哦!

可观测性指标

可观测性指标主要包括两个大类: 服务端指标和客户端指标, 服务端指标由服务端直接生成, 客户端指标在客户端产生, 由服务端通过 rpc 请求客户端获取到. 客户端指标又可细分为生产端指标和消费端指标.所有 87 个可观测性指标及其主要含义如下:

服务端指标

服务端指标

指标名称含义对应Broker指标名
rocketmq_broker_tpsBroker级别的生产TPS
rocketmq_broker_qpsBroker级别的消费QPS
rocketmq_broker_commitlog_diffBroker组从节点同步落后消息size
rocketmq_brokeruntime_pmdt_0ms服务端开始处理写请求到完成写入的耗时(0ms)putMessageDistributeTime
rocketmq_brokeruntime_pmdt_0to10ms服务端开始处理写请求到完成写入的耗时(0~10ms)
rocketmq_brokeruntime_pmdt_10to50ms服务端开始处理写请求到完成写入的耗时(10~50ms)
rocketmq_brokeruntime_pmdt_50to100ms服务端开始处理写请求到完成写入的耗时(50~100ms)
rocketmq_brokeruntime_pmdt_100to200ms服务端开始处理写请求到完成写入的耗时(100~200ms)
rocketmq_brokeruntime_pmdt_200to500ms服务端开始处理写请求到完成写入的耗时(200~500ms)
rocketmq_brokeruntime_pmdt_500to1s服务端开始处理写请求到完成写入的耗时(500~1000ms)
rocketmq_brokeruntime_pmdt_1to2s服务端开始处理写请求到完成写入的耗时(1~2s)
rocketmq_brokeruntime_pmdt_2to3s服务端开始处理写请求到完成写入的耗时(2~3s)
rocketmq_brokeruntime_pmdt_3to4s服务端开始处理写请求到完成写入的耗时(3~4s)
rocketmq_brokeruntime_pmdt_4to5s服务端开始处理写请求到完成写入的耗时(4~5s)
rocketmq_brokeruntime_pmdt_5to10s服务端开始处理写请求到完成写入的耗时(5~10s)
rocketmq_brokeruntime_pmdt_10stomore服务端开始处理写请求到完成写入的耗时(> 10s)
rocketmq_brokeruntime_dispatch_behind_bytes到现在为止,未被分发(构建索引之类的操作)的消息bytesdispatchBehindBytes
rocketmq_brokeruntime_put_message_size_totalbroker写入消息size的总和putMessageSizeTotal
rocketmq_brokeruntime_put_message_average_sizebroker写入消息的平均大小putMessageAverageSize
rocketmq_brokeruntime_remain_transientstore_buffer_numbsTransientStorePool 中队列的容量remainTransientStoreBufferNumbs
rocketmq_brokeruntime_earliest_message_timestampbroker存储的消息最早的时间戳earliestMessageTimeStamp
rocketmq_brokeruntime_putmessage_entire_time_maxbroker自运行以来,写入消息耗时的最大值putMessageEntireTimeMax
rocketmq_brokeruntime_start_accept_sendrequest_time开始接受发送请求的时间startAcceptSendRequestTimeStamp
rocketmq_brokeruntime_putmessage_times_totalbroker写入消息的总次数putMessageTimesTotal
rocketmq_brokeruntime_getmessage_entire_time_maxbroker自启动以来,处理消息拉取的最大耗时getMessageEntireTimeMax
rocketmq_brokeruntime_pagecache_lock_time_millspageCacheLockTimeMills
rocketmq_brokeruntime_commitlog_disk_ratiocommitLog所在磁盘的使用比例commitLogDiskRatio
rocketmq_brokeruntime_dispatch_maxbufferbroker没有计算,一直为0dispatchMaxBuffer
rocketmq_brokeruntime_pull_threadpoolqueue_capacity处理拉取请求线程池队列的容量pullThreadPoolQueueCapacity
rocketmq_brokeruntime_send_threadpoolqueue_capacity处理发送请求线程池队列的容量sendThreadPoolQueueCapacity
rocketmq_brokeruntime_query_threadpool_queue_capacity处理查询请求线程池队列的容量queryThreadPoolQueueCapacity
rocketmq_brokeruntime_pull_threadpoolqueue_size处理拉取请求线程池队列的实际sizepullThreadPoolQueueSize
rocketmq_brokeruntime_query_threadpoolqueue_size处理查询请求线程池队列的实际sizequeryThreadPoolQueueSize
rocketmq_brokeruntime_send_threadpool_queue_size处理send请求线程池队列的实际sizesendThreadPoolQueueSize
rocketmq_brokeruntime_pull_threadpoolqueue_headwait_timemills处理拉取请求线程池队列的队头任务等待时间pullThreadPoolQueueHeadWaitTimeMills
rocketmq_brokeruntime_query_threadpoolqueue_headwait_timemills处理查询请求线程池队列的队头任务等待时间queryThreadPoolQueueHeadWaitTimeMills
rocketmq_brokeruntime_send_threadpoolqueue_headwait_timemills处理发送请求线程池队列的队头任务等待时间sendThreadPoolQueueHeadWaitTimeMills
rocketmq_brokeruntime_msg_gettotal_yesterdaymorning到昨晚12点为止,读取消息的总次数msgGetTotalYesterdayMorning
rocketmq_brokeruntime_msg_puttotal_yesterdaymorning到昨晚12点为止,写入消息的总次数msgPutTotalYesterdayMorning
rocketmq_brokeruntime_msg_gettotal_todaymorning到今晚12点为止,读取消息的总次数msgGetTotalTodayMorning
rocketmq_brokeruntime_msg_puttotal_todaymorning到昨晚12点为止,写入消息的总次数putMessageTimesTotal
rocketmq_brokeruntime_msg_put_total_today_now每个broker到现在为止,写入的消息次数msgPutTotalTodayNow
rocketmq_brokeruntime_msg_gettotal_today_now每个broker到现在为止,读取的消息次数msgGetTotalTodayNow
rocketmq_brokeruntime_commitlogdir_capacity_freecommitLog所在目录的可用空间commitLogDirCapacity
rocketmq_brokeruntime_commitlogdir_capacity_totalcommitLog所在目录的总空间
rocketmq_brokeruntime_commitlog_maxoffsetcommitLog的最大offsetcommitLogMaxOffset
rocketmq_brokeruntime_commitlog_minoffsetcommitLog的最小offsetcommitLogMinOffset
rocketmq_brokeruntime_remain_howmanydata_toflushremainHowManyDataToFlush
rocketmq_brokeruntime_getfound_tps600600s内getMessage时get到消息的平均TPSgetFoundTps
rocketmq_brokeruntime_getfound_tps6060s内getMessage时get到消息的平均TPS
rocketmq_brokeruntime_getfound_tps1010s内getMessage时get到消息的平均TPS
rocketmq_brokeruntime_gettotal_tps600600s内getMessage次数的平均TPSgetTotalTps
rocketmq_brokeruntime_gettotal_tps6060s内getMessage次数的平均TPS
rocketmq_brokeruntime_gettotal_tps1010s内getMessage次数的平均TPS
rocketmq_brokeruntime_gettransfered_tps600getTransferedTps
rocketmq_brokeruntime_gettransfered_tps60
rocketmq_brokeruntime_gettransfered_tps10
rocketmq_brokeruntime_getmiss_tps600600s内getMessage时没有get到消息的平均TPSgetMissTps
rocketmq_brokeruntime_getmiss_tps6060s内getMessage时没有get到消息的平均TPS
rocketmq_brokeruntime_getmiss_tps1010s内getMessage时没有get到消息的平均TPS
rocketmq_brokeruntime_put_tps600600s内写入消息次数的平均TPSputTps
rocketmq_brokeruntime_put_tps6060s内写入消息次数的平均TPS
rocketmq_brokeruntime_put_tps1010s内写入消息次数的平均TPS
指标名称含义对应Broker指标名
rocketmq_broker_tpsBroker级别的生产TPS
rocketmq_broker_qpsBroker级别的消费QPS
rocketmq_broker_commitlog_diffBroker组从节点同步落后消息size
rocketmq_brokeruntime_pmdt_0ms服务端开始处理写请求到完成写入的耗时(0ms)putMessageDistributeTime
rocketmq_brokeruntime_pmdt_0to10ms服务端开始处理写请求到完成写入的耗时(0~10ms)
rocketmq_brokeruntime_pmdt_10to50ms服务端开始处理写请求到完成写入的耗时(10~50ms)
rocketmq_brokeruntime_pmdt_50to100ms服务端开始处理写请求到完成写入的耗时(50~100ms)
rocketmq_brokeruntime_pmdt_100to200ms服务端开始处理写请求到完成写入的耗时(100~200ms)
rocketmq_brokeruntime_pmdt_200to500ms服务端开始处理写请求到完成写入的耗时(200~500ms)
rocketmq_brokeruntime_pmdt_500to1s服务端开始处理写请求到完成写入的耗时(500~1000ms)
rocketmq_brokeruntime_pmdt_1to2s服务端开始处理写请求到完成写入的耗时(1~2s)
rocketmq_brokeruntime_pmdt_2to3s服务端开始处理写请求到完成写入的耗时(2~3s)
rocketmq_brokeruntime_pmdt_3to4s服务端开始处理写请求到完成写入的耗时(3~4s)
rocketmq_brokeruntime_pmdt_4to5s服务端开始处理写请求到完成写入的耗时(4~5s)
rocketmq_brokeruntime_pmdt_5to10s服务端开始处理写请求到完成写入的耗时(5~10s)
rocketmq_brokeruntime_pmdt_10stomore服务端开始处理写请求到完成写入的耗时(> 10s)
rocketmq_brokeruntime_dispatch_behind_bytes到现在为止,未被分发(构建索引之类的操作)的消息bytesdispatchBehindBytes
rocketmq_brokeruntime_put_message_size_totalbroker写入消息size的总和putMessageSizeTotal
rocketmq_brokeruntime_put_message_average_sizebroker写入消息的平均大小putMessageAverageSize
rocketmq_brokeruntime_remain_transientstore_buffer_numbsTransientStorePool 中队列的容量remainTransientStoreBufferNumbs
rocketmq_brokeruntime_earliest_message_timestampbroker存储的消息最早的时间戳earliestMessageTimeStamp
rocketmq_brokeruntime_putmessage_entire_time_maxbroker自运行以来,写入消息耗时的最大值putMessageEntireTimeMax
rocketmq_brokeruntime_start_accept_sendrequest_time开始接受发送请求的时间startAcceptSendRequestTimeStamp
rocketmq_brokeruntime_putmessage_times_totalbroker写入消息的总次数putMessageTimesTotal
rocketmq_brokeruntime_getmessage_entire_time_maxbroker自启动以来,处理消息拉取的最大耗时getMessageEntireTimeMax
rocketmq_brokeruntime_pagecache_lock_time_millspageCacheLockTimeMills
rocketmq_brokeruntime_commitlog_disk_ratiocommitLog所在磁盘的使用比例commitLogDiskRatio
rocketmq_brokeruntime_dispatch_maxbufferbroker没有计算,一直为0dispatchMaxBuffer
rocketmq_brokeruntime_pull_threadpoolqueue_capacity处理拉取请求线程池队列的容量pullThreadPoolQueueCapacity
rocketmq_brokeruntime_send_threadpoolqueue_capacity处理发送请求线程池队列的容量sendThreadPoolQueueCapacity
rocketmq_brokeruntime_query_threadpool_queue_capacity处理查询请求线程池队列的容量queryThreadPoolQueueCapacity
rocketmq_brokeruntime_pull_threadpoolqueue_size处理拉取请求线程池队列的实际sizepullThreadPoolQueueSize
rocketmq_brokeruntime_query_threadpoolqueue_size处理查询请求线程池队列的实际sizequeryThreadPoolQueueSize
rocketmq_brokeruntime_send_threadpool_queue_size处理send请求线程池队列的实际sizesendThreadPoolQueueSize
rocketmq_brokeruntime_pull_threadpoolqueue_headwait_timemills处理拉取请求线程池队列的队头任务等待时间pullThreadPoolQueueHeadWaitTimeMills
rocketmq_brokeruntime_query_threadpoolqueue_headwait_timemills处理查询请求线程池队列的队头任务等待时间queryThreadPoolQueueHeadWaitTimeMills
rocketmq_brokeruntime_send_threadpoolqueue_headwait_timemills处理发送请求线程池队列的队头任务等待时间sendThreadPoolQueueHeadWaitTimeMills
rocketmq_brokeruntime_msg_gettotal_yesterdaymorning到昨晚12点为止,读取消息的总次数msgGetTotalYesterdayMorning
rocketmq_brokeruntime_msg_puttotal_yesterdaymorning到昨晚12点为止,写入消息的总次数msgPutTotalYesterdayMorning
rocketmq_brokeruntime_msg_gettotal_todaymorning到今晚12点为止,读取消息的总次数msgGetTotalTodayMorning
rocketmq_brokeruntime_msg_puttotal_todaymorning到昨晚12点为止,写入消息的总次数putMessageTimesTotal
rocketmq_brokeruntime_msg_put_total_today_now每个broker到现在为止,写入的消息次数msgPutTotalTodayNow
rocketmq_brokeruntime_msg_gettotal_today_now每个broker到现在为止,读取的消息次数msgGetTotalTodayNow
rocketmq_brokeruntime_commitlogdir_capacity_freecommitLog所在目录的可用空间commitLogDirCapacity
rocketmq_brokeruntime_commitlogdir_capacity_totalcommitLog所在目录的总空间
rocketmq_brokeruntime_commitlog_maxoffsetcommitLog的最大offsetcommitLogMaxOffset
rocketmq_brokeruntime_commitlog_minoffsetcommitLog的最小offsetcommitLogMinOffset
rocketmq_brokeruntime_remain_howmanydata_toflushremainHowManyDataToFlush
rocketmq_brokeruntime_getfound_tps600600s内getMessage时get到消息的平均TPSgetFoundTps
rocketmq_brokeruntime_getfound_tps6060s内getMessage时get到消息的平均TPS
rocketmq_brokeruntime_getfound_tps1010s内getMessage时get到消息的平均TPS
rocketmq_brokeruntime_gettotal_tps600600s内getMessage次数的平均TPSgetTotalTps
rocketmq_brokeruntime_gettotal_tps6060s内getMessage次数的平均TPS
rocketmq_brokeruntime_gettotal_tps1010s内getMessage次数的平均TPS
rocketmq_brokeruntime_gettransfered_tps600getTransferedTps
rocketmq_brokeruntime_gettransfered_tps60
rocketmq_brokeruntime_gettransfered_tps10
rocketmq_brokeruntime_getmiss_tps600600s内getMessage时没有get到消息的平均TPSgetMissTps
rocketmq_brokeruntime_getmiss_tps6060s内getMessage时没有get到消息的平均TPS
rocketmq_brokeruntime_getmiss_tps1010s内getMessage时没有get到消息的平均TPS
rocketmq_brokeruntime_put_tps600600s内写入消息次数的平均TPSputTps
rocketmq_brokeruntime_put_tps6060s内写入消息次数的平均TPS
rocketmq_brokeruntime_put_tps1010s内写入消息次数的平均TPS
生产端指标

生产端指标

指标名称含义
rocketmq_producer_offsettopic当前时间的最大offset
rocketmq_topic_retry_offset重试Topic当前时间的最大offset
rocketmq_topic_dlq_offset死信Topic当前时间的最大offset
rocketmq_producer_tpsTopic在一个Broker组上的生产TPS
rocketmq_producer_message_sizeTopic在一个Broker组上的生产消息大小的TPS
rocketmq_queue_producer_tps队列级别生产TPS
rocketmq_queue_producer_message_size队列级别生产消息大小的TPS
消费端指标
### 消费端指标
指标名称含义
rocketmq_group_diff消费组消息堆积消息数
rocketmq_group_retrydiff消费组重试队列堆积消息数
rocketmq_group_dlqdiff消费组死信队列堆积消息数
rocketmq_group_count消费组内消费者个数
rocketmq_client_consume_fail_msg_count过去1h消费者消费失败的次数
rocketmq_client_consume_fail_msg_tps消费者消费失败的TPS
rocketmq_client_consume_ok_msg_tps消费者消费成功的TPS
rocketmq_client_consume_rt消息从拉取到被消费的时间
rocketmq_client_consumer_pull_rt客户端拉取消息的时间
rocketmq_client_consumer_pull_tps客户端拉取消息的TPS
rocketmq_consumer_tps每个Broker组上订阅组的消费TPS
rocketmq_group_consume_tps订阅组当前消费TPS(对rocketmq_consumer_tps按broker聚合)
rocketmq_consumer_offset订阅组在一个broker组上当前的消费Offset
rocketmq_group_consume_total_offset订阅组当前消费的Offset(对rocketmq_consumer_offset按broker聚合)
rocketmq_consumer_message_size订阅组在一个broker组上消费消息大小的TPS
rocketmq_send_back_nums订阅组在一个broker组上消费失败,写入重试消息的次数
rocketmq_group_get_latency_by_storetime消费组消费延时,exporter get到消息后与当前时间相减