Recommended nvme_core.io_timeout value in Azure

This document (000021813) is provided subject to the disclaimer at the end of this document.

Environment

SUSE Linux Enterprise Server
Microsoft Azure Virtual Machines

Situation

What is the recommended value of nvme_core.io_timeout in the Azure cloud?

Resolution

Microsoft Engineering recommends setting nvme_core.io_timeout to 240 seconds.

To add this parameter:

Open the /etc/sysconfig/grub configuration file with an editor
Modify the GRUB_CMDLINE_LINUX_DEFAULT parameter adding nvme_core.io_timeout=240 to the existing string value
Regenerate GRUB2 with the following command:
```
# grub2-mkconfig -o /boot/grub2/grub.cfg
```
Reboot
Check the kernel command line and sysfs to confirm the parameter is set with the following commands:
```
# cat /proc/cmdline
# cat /sys/module/nvme_core/parameters/io_timeout
```

Status

Reported to Engineering

Additional Information

The default io_timeout parameter value for the nvme-core kernel module is 30 seconds. In the Azure cloud, the virtualization host considers a timeout of 120 seconds to be an NVMe failure. Microsoft prefers that the OS not take any action in the event of NVMe failure but to allow the virtualization host to detect and take action on VM recovery.

Please contact Azure support if a newly published image does not have this setting on the kernel command line.

Disclaimer

This Support Knowledgebase provides a valuable tool for SUSE customers and parties interested in our products and solutions to acquire information, ideas and learn from one another. Materials are provided for informational, personal or non-commercial use within your organization and are presented "AS IS" WITHOUT WARRANTY OF ANY KIND.