NAME

diag_nvme - diagnose NVMe devices

SYNOPSIS

diag_nvme [<nvmen ...>]

diag_nvme [-d <file>] [-f <file>] <nvmen>

diag_nvme --help

DESCRIPTION

The diag_nvme command retrieves the SMART data from the NVMe device(s) specified and report failures detected through events created in the servicelog database. If no device is specified, all NVMe devices detected in the system will go through the diagnostics procedure.

The user can control which events will be reported through the configuration file /etc/ppc64-diag/diag_nvme.config

OPTIONS

-d, --dump file

Dump SMART data to the specified path and file name file. The SMART data is extracted from an NVMe device, so specifying one is mandatory if this option is selected. File created is in a simple key=value format.

-f, --file file

This option usage is for testing only. Use SMART data from the specified path and file name file instead of device, one NVMe is mandatory if this option is selected. The expected format of the file is a simple key=value that is the same one provided with the -d / --dump option. If file is missing from the filesystem it will be treated as a failure to retrieve SMART data and an event will be reported.

-h, --help

Print a help message and exit

EXAMPLES

diag_nvme

Run diagnostics in all NVMe devices detected in the system.

diag_nvme nvme0 nvme1

Run diagnostics only in nvme0 and nvme1 devices.

diag_nvme -d smart.txt nvme0

Dump SMART data from nvme0 into file smart.txt.

diag_nvme -f smart.txt nvme0

Read SMART data from file smart.txt and use it as health information for diagnostics of device nvme0.

REPORTING BUGS

Patches and issues may be submitted at https://github.com/power-ras/ppc64-diag/

FILES

/etc/ppc64-diag/diag_nvme.config

Configuration file to select which events to report to the servicelog database.

/etc/cron.daily/run_diag_nvme

Script to run diag_nvme in the system on a daily basis.

SEE ALSO

servicelog(8)

AUTHORS

Murilo Fossa Vicentini