DocsICP FilingConsole
官方文档
首页

Analyze application exceptions by using coredump

更新时间：
复制 MD 格式
产品详情
我的收藏

A coredump is a file containing an application's memory state when it crashes or terminates unexpectedly. The operating system creates this file for later debugging and analysis. This topic describes how to enable coredump for an Elastic Container Instance (ECI) instance. This allows you to analyze core files from unexpected container terminations to identify the cause of a problem and fix the application.

Background information

On Linux, when an application terminates unexpectedly or crashes, the operating system records the application's memory state and saves it to a file. This process is known as a coredump. You can then analyze the core file to identify the cause of the problem.

The following figure shows the Linux signals that trigger a coredump. For these signals, the default action is Core.

For more information, see coredump file.

Overview

By default, coredump is disabled for an ECI instance to prevent excessive disk usage, which could lead to service outages. You can enable coredump by using one of the following methods:

  • Method 1: Enable a coredump O&M task

    After you manually enable coredump, the system creates an O&M task. When a container terminates abnormally or exits, the generated core file is automatically saved to OSS.

  • Method 2: Customize the save path for core files

    You can store core files in external storage by specifying a custom save path. After you set the save path, coredump is automatically enabled. When a container crashes or exits unexpectedly, the resulting core file is saved to the specified path.

Note

  • Method 1 is easy to use but has time and region limitations. It is suitable for temporary debugging and diagnostics.

    • The generated O&M task is a one-time task. Once a core file is successfully captured, coredump is automatically disabled. The O&M task is valid for 12 hours and expires if it is not triggered within this period.

    • This method is not supported in the following regions: China (Ulanqab), China (Heyuan), China (Guangzhou), China (Nanjing), Philippines (Manila), South Korea (Seoul), and Thailand (Bangkok).

  • Method 2 requires you to configure external storage. If your application is unstable, you can use this method to ensure that core files are captured. However, if a faulty application restarts repeatedly, a large number of core files can be generated.

Method 1: Enable a coredump O&M task

Console

  1. Log on to the Elastic Container Instance console.

  2. Enable coredump for the ECI instance.

    1. Click the ID of the target instance to open the instance details page.

    2. Click the O&M tab, select the Coredump tab, and then click Enable.

      After you enable coredump, the system generates an O&M task. If coredump has not been triggered, the task status is Pending.

      coredump-1

  3. Trigger a coredump.

    Connect to the ECI instance. In the container, run the sleep 100 command and then press Ctrl+\ to trigger a coredump. The generated core file is automatically saved to OSS.coredump2

  4. Download the core file.

    After the coredump is triggered and the core file is generated, the status of the O&M task changes to Successful. Click Download in the Result column to download the core file.

    coredump-2

    Note

    If the download does not start, check the site permission settings in your browser.

API

  1. Enable coredump for the ECI instance.

    Call the CreateInstanceOpsTask operation to create an O&M task. Specify the target ECI instance, set the OpsType parameter to coredump, and set the OpsValue parameter to enable. For more information, see CreateInstanceOpsTask.

    Note

    When you specify the ECI instance, make sure that the CorePattern parameter was not set during instance creation.

  2. Trigger a coredump.

    Connect to the ECI instance. In the container, run the sleep 100 command and then press Ctrl+\ to trigger a coredump. The generated core file is automatically saved to OSS.coredump2

  3. Download the core file.

    Call the DescribeInstanceOpsRecords operation to view the result of the O&M task. In the response, the ResultContent field contains the URL of the core file in OSS. You can access the URL to download the core file.

Method 2: Customize the save path

Core files are typically used for offline analysis. Therefore, we recommend that you save them to external storage instead of a local path within the container. This prevents the core files from being lost when the container exits. ECI lets you specify a custom save path for core files. After you set the path, coredump is automatically enabled.

When you call the CreateContainerGroup operation to create an ECI instance, you can pass the CorePattern parameter to set the save path for core files. The following table describes the related parameters. For more information, see CreateContainerGroup.

Note

Configure the volume parameters according to your volume type.

Parameter

Type

Example

Description

CorePattern

String

/data/dump-a/core

The save path for core files. This path must be within a mounted volume.

Important

The path cannot start with a pipe character (|). This means you cannot configure an executable program using coredump.

Volume.N.Name

String

volume1

The name of the volume.

Volume.N.Type

String

NFSVolume

The type of the volume. In this example, NFSVolume is used.

Volume.N.NFSVolume.Path

String

/dump

The path of the NFS volume.

Volume.N.NFSVolume.Server

String

143b24****-gfn3.cn-beijing.nas.aliyuncs.com

The address of the NFS server.

If you use Alibaba Cloud File Storage NAS, this is the address of the mount point of the NAS file system.

Container.N.VolumeMount.N.Name

String

volume1

The name of the volume to mount. This must match the value of Volume.N.Name.

Container.N.VolumeMount.N.MountPath

String

/data/dump-a/

The mount path.

The volume's content overwrites any existing content at the mount path within the container. Make sure that the path is correct.

The following example shows how to use an Alibaba Cloud File Storage NAS file system as external storage:

  1. Create an ECI instance called instance A, mount the NAS file system, and set the save path for core files.

    Call the CreateContainerGroup operation to create instance A. Set the following parameters to mount the /dump/ directory from the NAS file system to /data/dump-a/ in the container, and set the core file save path to /data/dump-a/core.

    ContainerGroupName=test-a
SecurityGroupId=sg-bp1daxpbz9lzpvvc****
VSwitchId=vsw-bp1gds63lmlm7ib05****
# Declare a volume
Volume.1.Name=volume1
Volume.1.Type=NFSVolume
Volume.1.NFSVolume.Path=/dump/
Volume.1.NFSVolume.Server=143b24****-gfn3.cn-beijing.nas.aliyuncs.com
# Mount the volume to the container
Container.1.Name=nginx
Container.1.Image=registry-vpc.cn-beijing.aliyuncs.com/eci_open/nginx:1.14.2
Container.1.VolumeMount.1.Name=volume1
Container.1.VolumeMount.1.MountPath=/data/dump-a/
# Set the save path for core files
CorePattern=/data/dump-a/core

  2. Trigger a coredump in any directory in instance A's container.

    In the container, run the sleep 100 command and then press Ctrl+\ to trigger a coredump. The core file is saved to /data/dump-a/core in the container.

    coredump4

  3. Release instance A.

  4. Mount the same NAS file system to another ECI instance called instance B.

    Call the CreateContainerGroup operation to create instance B. Pass the following parameters to mount the same /dump/ directory of the NAS file system to the /data/dump-b/ directory in the container.

    ContainerGroupName=test-nas-b
SecurityGroupId=sg-bp1daxpbz9lzpvvc****
VSwitchId=vsw-bp1gds63lmlm7ib05****
# Declare a volume
Volume.1.Name=volume1
Volume.1.Type=NFSVolume
Volume.1.NFSVolume.Path=/dump/
Volume.1.NFSVolume.Server=143b24****-gfn3.cn-beijing.nas.aliyuncs.com
# Mount the volume to the container
Container.1.Name=nginx
Container.1.Image=registry-vpc.cn-beijing.aliyuncs.com/eci_open/nginx:1.14.2
Container.1.VolumeMount.1.Name=volume1
Container.1.VolumeMount.1.MountPath=/data/dump-b/

  5. View the core file in instance B's container.

    As shown in the following figure, you can view the core file in the /data/dump-b directory of the container. Because the core file is saved to external storage, it is not lost after instance A is released and remains available for analysis.

    coredump5

FAQ

The download does not start when I click Download in the console. What should I do?

If the download does not start, check the site permission settings in your browser. For example, in Chrome, follow these steps to grant the necessary permission:

  1. Open the Elastic Container Instance console, click the 浏览器图标 icon to the left of the browser address bar, and then select Site settings.浏览器设置1

  2. Set Insecure content to Allow.浏览器设置2

Previous：OperationsNext：Analyze network issues with tcpdump