Setting the kernel.hung_task_panic tunable

Last published : Jun 11, 2026
The topic applies to SFHA, SFCFSHA, and VCS.
By default, in the Linux kernel the kernel.hung_task_panic tunable is enabled and the kernel.hung_task_timeout_secs tunable is set to a default non-zero value.
To ensure that the node does not panic, the kernel.hung_task_panic tunable must be disabled. If kernel.hung_task_panic is enabled, then it causes the kernel to panic when any of the following kernel threads waits for more than the kernel.hung_task_timeout_secs value:
  • The vxfenconfig thread in the vxfen configuration path waits for GAB to seed.
  • The vxfenswap thread in the online coordinator disks replacement path waits for the snapshot of peer nodes of the new coordinator disks.
To disable the kernel.hung_task_panic tunable:
  • Set the kernel.hung_task_panic tunable to zero (0) in the /etc/sysctl.conf file. This step ensures that the change is persistent across node restarts.
  • Run the command on each node.
Related information

sysctl -w kernel.hung_task_panic=0

To verify the kernel.hung_task_panic tunable value, run the following command:
  • \# sysctl -a | grep hung_task_panic