tencent cloud

Feedback

Kernel and IO Issues

Last updated: 2024-01-06 17:32:18
    The status check provides a report of instance exceptions. This document mainly describes the symptoms, causes and solutions of kernel and IO problems shown in the status check report.

    Troubleshooting Kernel Failures

    Problems

    The kernel failure may cause login failure or abnormal restart.

    Common causes

    Kernel hung_task

    The kernel hung task is based on a single kernel thread named as khungtaskd, which monitors processes in the TASK_UNINTERRUPTIBLE status. If a process stuck in D state during the period specified by kernel.hung_task_timeout_secs (defaults to 120 seconds), the stack information of this hung task process will be printed.
    If kernel.hung_task_panic=1 is configured, the hung task will trigger kernel panic and system restart.

    Kernel soft lockup

    A soft lockup refers to a kernel thread using and not releasing a CPU, without giving other tasks a chance to run. Each CPU is assigned with a timed kernel thread watchdog/x. If this thread is not executed during the specified period (the default period is two times the kernel.watchdog_thresh value. For example, the default kernel.watchdog_thresh value is 10 seconds for a 3.10 kernel), soft lockup occurs.
    If kernel.softlockup_panic=1 is configured, the soft lockup will trigger kernel panic and system restart.

    Kernel panic

    A kernel panic refers to a kernel crash that causes the abnormal restart. The kernel panic will be generally caused by:
    Kernel hung_task, with kernel.hung_task_panic=1 configured.
    Kernel soft lockup, with kernel.softlockup_panic=1 configured.
    Kernel bug

    Solution

    Due to the difficulty, we recommend you submit a ticket for the troubleshooting.

    Troubleshooting Disk Failures

    The inode is full

    Problem: the error message “No space left on device” is prompted when you create a file. After you run the df -i command, you will see inode is 100% used. Common causes: the file system exhausted all inodes. Procedure: delete useless files or expand the disk.

    The disk space is full

    Problem: the error message “No space left on device” is prompted when you create a file. After you run the df -h command, you will see the disk space is 100% used. Common causes: the disk space runs out. Procedure: delete useless files or expand the disk.

    The disk is read-only

    Problem: the file system can read files only without creating one. Common cause: the file system is damaged. Procedure:
    1. Create a snapshot to back up the disk data. For detailed directions, see Creating Snapshots.
    2. Perform the troubleshooting procedure according to the disk type.
    System disk
    Data disk
    We recommend directly restarting the instance, please see Restart Instances.
    1. Run the following command to check the type of the read-only disk file system.
    lsblk -f
    2. Run the following command to detach the data disk.
    umount <mount point of the data disk>
    3. Run the file system-specific command to fix the file system.
    Run the following command on the ext3/ext4 file system:
    fsck -y /dev/[data disk]
    Run the following command on the xfs file system:
    xfs_repair /dev/[data disk]

    The disk %util is high

    Problem: the instance lags, and responds slowly or stop responding to the SSH or VNC login. Common cause: high IO causes the disk %util to reach 100%. Procedure: check the high IO status, and assess whether to reduce IO reads/writes, or use a disk with higher performance.
    Contact Us

    Contact our sales team or business advisors to help your business.

    Technical Support

    Open a ticket if you're looking for further assistance. Our Ticket is 7x24 avaliable.

    7x24 Phone Support