Help Center/ Cloud Container Engine/ FAQs/ Workload/ Workload Abnormalities/ What Should I Do If Container Startup Fails?
Updated on 2024-09-04 GMT+08:00

What Should I Do If Container Startup Fails?

Fault Locating

On the details page of a workload, if an event is displayed indicating that the container fails to be started, perform the following steps to locate the fault:

  1. Log in to the node where the abnormal workload is located.
  2. Check the ID of the container where the workload pod exits abnormally.

    docker ps -a | grep $podName

  3. View the logs of the corresponding container.

    docker logs $containerID

    Rectify the fault of the workload based on logs.

  4. Check the error logs.

    cat /var/log/messages | grep $containerID  | grep oom

    Check whether the system OOM is triggered based on the logs.

Troubleshooting Process

Determine the cause based on the event information, as listed in Table 1.

Table 1 Container startup failure

Log or Event

Cause and Solution

The log contains exit(0).

No process exists in the container.

Check whether the container is running properly.

Check Item 1: Whether There Are Processes that Keep Running in the Container (Exit Code: 0)

Event information: Liveness probe failed: Get http...

The log contains exit(137).

Health check fails.

Check Item 2: Whether Health Check Fails to Be Performed (Exit Code: 137)

Event information:

Thin Pool has 15991 free data blocks which are less than minimum required 16383 free data blocks. Create more free space in thin pool or use dm.min_free_space option to change behavior

The disk space is insufficient. Clear the disk space.

Check Item 3: Whether the Container Disk Space Is Insufficient

The keyword OOM exists in the log.

The memory is insufficient.

Check Item 4: Whether the Upper Limit of Container Resources Has Been Reached

Check Item 5: Whether the Resource Limits Are Improperly Configured for the Container

Address already in use

A conflict occurs between container ports in the pod.

Check Item 6: Whether the Container Ports in the Same Pod Conflict with Each Other

Error: failed to start container "filebeat": Error response from daemon: OCI runtime create failed: container_linux.go:330: starting container process caused "process_linux.go:381: container init caused \"setenv: invalid argument\"": unknown

A secret is mounted to the workload, and the value of the secret is not encrypted using Base64.

Check Item 7: Whether the Value of the Secret Mounted to the Workload Meets Requirements

In addition to the preceding possible causes, there are some other possible causes:

Figure 1 Troubleshooting process of the container restart failure

Check Item 1: Whether There Are Processes that Keep Running in the Container (Exit Code: 0)

  1. Log in to the node where the abnormal workload is located.
  2. View the container status.

    docker ps -a | grep $podName

    Example:

    If no running process exists in the container, the status code Exited (0) is displayed.

Check Item 2: Whether Health Check Fails to Be Performed (Exit Code: 137)

The health check configured for a workload is performed on services periodically. If an exception occurs, the pod reports an event and the pod fails to be restarted.

If the liveness-type (workload liveness probe) health check is configured for the workload and the number of health check failures exceeds the threshold, the containers in the pod will be restarted. On the workload details page, if Kubernetes events contain Liveness probe failed: Get http..., the health check fails.

Solution

Click the workload name to go to the workload details page, click the Containers tab. Then select Health Check to check whether the policy is proper or whether services are running properly.

Check Item 3: Whether the Container Disk Space Is Insufficient

The following message refers to the thin pool disk that is allocated from the Docker disk selected during node creation. You can run the lvs command as user root to view the current disk usage.

Thin Pool has 15991 free data blocks which are less than minimum required 16383 free data blocks. Create more free space in thin pool or use dm.min_free_space option to change behavior

Solution

Solution 1: Clearing images

Perform the following operations to clear unused images:
  • Nodes that use containerd
    1. Obtain local images on the node.
      crictl images -v
    2. Delete the images that are not required by image ID.
      crictl rmi Image ID
  • Nodes that use Docker
    1. Obtain local images on the node.
      docker images
    2. Delete the images that are not required by image ID.
      docker rmi Image ID

Do not delete system images such as the cce-pause image. Otherwise, pods may fail to be created.

Solution 2: Expanding the disk capacity

To expand a disk capacity, perform the following steps:

  1. Expand the capacity of a data disk on the EVS console. For details, see Expanding EVS Disk Capacity.

    Only the storage capacity of the EVS disk is expanded. You also need to perform the following steps to expand the capacity of the logical volume and file system.

  2. Log in to the CCE console and click the cluster. In the navigation pane, choose Nodes. Click More > Sync Server Data in the row containing the target node.
  3. Log in to the target node.
  4. Run the lsblk command to check the block device information of the node.

    A data disk is divided depending on the container storage Rootfs:

    Overlayfs: No independent thin pool is allocated. Image data is stored in dockersys.

    1. Check the disk and partition sizes of the device.
      # lsblk
      NAME                MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
      sda                   8:0    0   50G  0 disk 
      └─sda1                8:1    0   50G  0 part /
      sdb                   8:16   0  150G  0 disk      # The data disk has been expanded to 150 GiB, but 50 GiB space is not allocated.
      ├─vgpaas-dockersys  253:0    0   90G  0 lvm  /var/lib/containerd
      └─vgpaas-kubernetes 253:1    0   10G  0 lvm  /mnt/paas/kubernetes/kubelet
    2. Expand the disk capacity.

      Add the new disk capacity to the dockersys logical volume used by the container engine.

      1. Expand the PV capacity so that LVM can identify the new EVS capacity. /dev/sdb specifies the physical volume where dockersys is located.
        pvresize /dev/sdb

        Information similar to the following is displayed:

        Physical volume "/dev/sdb" changed
        1 physical volume(s) resized or updated / 0 physical volume(s) not resized
      2. Expand 100% of the free capacity to the logical volume. vgpaas/dockersys specifies the logical volume used by the container engine.
        lvextend -l+100%FREE -n vgpaas/dockersys

        Information similar to the following is displayed:

        Size of logical volume vgpaas/dockersys changed from <90.00 GiB (23039 extents) to 140.00 GiB (35840 extents).
        Logical volume vgpaas/dockersys successfully resized.
      3. Adjust the size of the file system. /dev/vgpaas/dockersys specifies the file system path of the container engine.
        resize2fs /dev/vgpaas/dockersys

        Information similar to the following is displayed:

        Filesystem at /dev/vgpaas/dockersys is mounted on /var/lib/containerd; on-line resizing required
        old_desc_blocks = 12, new_desc_blocks = 18
        The filesystem on /dev/vgpaas/dockersys is now 36700160 blocks long.
    3. Check whether the capacity is expanded.
      # lsblk
      NAME                MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
      sda                   8:0    0   50G  0 disk 
      └─sda1                8:1    0   50G  0 part /
      sdb                   8:16   0  150G  0 disk
      ├─vgpaas-dockersys  253:0    0   140G  0 lvm  /var/lib/containerd
      └─vgpaas-kubernetes 253:1    0   10G  0 lvm  /mnt/paas/kubernetes/kubelet

Check Item 4: Whether the Upper Limit of Container Resources Has Been Reached

If the upper limit of container resources has been reached, OOM will be displayed in the event details as well as in the log:

cat /var/log/messages | grep 96feb0a425d6 | grep oom

When a workload is created, if the requested resources exceed the configured upper limit, the system OOM is triggered and the container exits unexpectedly.

Check Item 5: Whether the Resource Limits Are Improperly Configured for the Container

If the resource limits set for the container during workload creation are less than required, the container fails to be restarted.

Check Item 6: Whether the Container Ports in the Same Pod Conflict with Each Other

  1. Log in to the node where the abnormal workload is located.
  2. Check the ID of the container where the workload pod exits abnormally.

    docker ps -a | grep $podName

  3. View the logs of the corresponding container.

    docker logs $containerID

    Rectify the fault of the workload based on logs. As shown in the following figure, container ports in the same pod conflict. As a result, the container fails to be started.

    Figure 2 Container restart failure due to a container port conflict

Solution

Re-create the workload and set a port number that is not used by any other pod.

Check Item 7: Whether the Value of the Secret Mounted to the Workload Meets Requirements

Information similar to the following is displayed in the event:

Error: failed to start container "filebeat": Error response from daemon: OCI runtime create failed: container_linux.go:330: starting container process caused "process_linux.go:381: container init caused \"setenv: invalid argument\"": unknown

The root cause is that a secret is mounted to the workload, but the value of the secret is not encrypted using Base64.

Solution:

Create a secret on the console. The value of the secret is automatically encrypted using Base64.

If you use YAML to create a secret, you need to manually encrypt its value using Base64.

# echo -n "Content to be encoded" | base64

Check Item 8: Whether the Container Startup Command Is Correctly Configured

The error messages are as follows:

Solution

Click the workload name to go to the workload details page, click the Containers tab. Choose Lifecycle, click Startup Command, and ensure that the command is correct.

Check Item 9: Whether the Java Probe Version Is latest

Kubernetes event "Created container init-pinpoint" occurs.

Solution

  1. When creating a workload, select the specific latest Java probe version (for example, 1.0.36, not the latest option) on the APM Settings tab in the Advanced Settings area.

  2. If you selected latest for the Java probe during workload creation, you can upgrade the workload and change it to the specific latest version (for example, 1.0.36).

Check Item 10: Whether the User Service Has a Bug

Check whether the workload startup command is correctly executed or whether the workload has a bug.

  1. Log in to the node where the abnormal workload is located.
  2. Check the ID of the container where the workload pod exits abnormally.

    docker ps -a | grep $podName

  3. View the logs of the corresponding container.

    docker logs $containerID

    Note: In the preceding command, containerID indicates the ID of the container that has exited.

    Figure 3 Incorrect startup command of the container

    As shown in the figure above, the container fails to be started due to an incorrect startup command. For other errors, rectify the bugs based on the logs.

Solution

Create a new workload and configure a correct startup command.