You need to enable JavaScript to run this app.
文档中心
E-MapReduce

E-MapReduce

复制全文
下载 pdf
EMR Serverless 队列最佳实践
EMR Serverless 接入日志服务TLS
复制全文
下载 pdf
EMR Serverless 接入日志服务TLS

EMR Serverless 支持对接火山引擎日志服务(TLS)产品,将作业相关操作的日志数据投递至 TLS 进行统一的加工、管理。本文为您介绍 EMR Serverless 对接 TLS 进行日志数据投递的操作要点。

使用限制

当前RayJob仅支持投递 Driver 日志,Spark 支持投递Driver、Executor日志,其他任务日志请在作业实例详情中查看。

前置配置
  1. 开通 TLS,并创建项目与日志主题。
  2. 完成 TLS 侧配置后,您可提交工单,联系火山引擎技术支持人员并提供TopicID,帮助您完成 EMR 侧的配置。当技术支持人员完成 EMR 侧配置后,后续即可将支持的 EMR Serverless 作业日志数据投递至您的 TLS 日志主题中。

查看日志详情

前置操作完成后,后续即可在 TLS 控制台中对应的项目和主题内查看详细的日志详情。

说明

前往 TLS 控制台查看日志详情时,需确保您的操作账号至少有 TLS 控制台的日志项目、日志详情的查看权限。

支持的日志类型及实例内容如下:

日志类型

示例

Spark Driver日志

{
    "__container_name__": "spark-kubernetes-driver",
    "__content__": "26/04/29 16:31:19 WARN [main] SparkConf: The configuration key 'spark.scheduler.listenerbus.eventqueue.size' has been deprecated as of Spark 2.3 and may be removed in the future. Please use the new key 'spark.scheduler.listenerbus.eventqueue.capacity' instead.",
    "__context_flow__": "1777451480893#c969921b8ec7b233-11690e6******1-0",
    "__image_name__": "emr-serverless-online-cn-beijing.cr.volces.com/emr-serverless-engine/spark:3.5.1-py3.12-ubuntu20.04-742",
    "__namespace__": "ns-2******-144000******",
    "__package_offset__": "1696903452055484",
    "__path__": "/var/data/spark-bda2526c-2e42-4c1b-a052-af8a7adb57b5/logs/syslog",
    "__pod_name__": "bq-3******-driver",
    "__pod_uid__": "48941247-2e26-45fe-bcd0-fd******",
    "__tag____receive_time__": "1777451481889",
    "__tag__account_id__": "emr_serverless_2******",
    "__time__": 1777451480893,
    "filename": "syslog"
}

Spark Executor日志

{
    "__content__": "I20260429 16:59:44.233684 140043121325824 ShuffleWriterNode.cc:144] init: forceShuffleWriterType = 0, preAllocSize 30 <= 2000, partitions = 1000, partitioning = 2, use VeloxShuffleWriterV2",
    "__context_flow__": "1777453185223#afb80dc061713c6b-af60ee******c-239",
    "__package_offset__": "30977949650294357",
    "__path__": "/data/ns-2******-1341100243457933312/bq-332652190-exec-7/var/data/spark-25d32af9-9d66-4ef8-97d6-484528d59798/logs/stderr",
    "__tag____receive_time__": "1777453185338",
    "__time__": 1777453185223,
    "account_id": "2******",
    "cluster_id": "ccvm01a30mbf393fuur0g",
    "filename": "stderr",
    "pod_name": "bq-332652190-exec-7"
}

RayJob 日志

{
    "__container_name__": "ray-head",
    "__content__": "\u001b[36m(UDFActor pid=5071, ip=9.127.161.106)\u001b[0m 2026-04-29 16:33:20 - PdfParseUDF-140438855291600 - WARNING - Got HTTP 429 (Too Many Requests), retrying (attempt 1/4) after 5.0s delay\u001b[32m [repeated 34x across cluster]\u001b[0m",
    "__context_flow__": "1777451603065#0a48b3a926d4a57b-7eef4c45df7d94f5-13856",
    "__image_name__": "las-ai-online-cn-beijing.cr.volces.com/las/ve-daft:0.7.2.post4-py3.11-ubuntu24.04",
    "__namespace__": "ns-2******-1441487549972348928",
    "__package_offset__": "2822841226614695",
    "__path__": "/tmp/ray/session_2026-04-28_22-34-19_670047_1/logs/job-driver-rayjob-3******-trntt.log",
    "__pod_name__": "rayjob-3******-nfp44-head-7lzth",
    "__pod_uid__": "73c3d104-5db1-4be3-99f3-76139c58c82a",
    "__tag____receive_time__": "1777451604567",
    "__tag__ray_job_name__": "rayjob-3******",
    "__time__": 1777451603065
}

常用查询语句

日志投递仅将日志投递到您的topic中,查询需要自行配置索引,可以根据tls提示从原始数据中提取索引字段。
常用查询语句如下:

  • 根据作业id查询
    • spark driver,如:pod_name : "bq-328853200-driver"
    • spark executor,如: pod_name: "bq-328853200-exec-7"
    • rayjob,如: tag__ray_job_name : "rayjob-328853201"
最近更新时间:2026.04.30 11:24:19
这个页面对您有帮助吗?
有用
有用
无用
无用