Logical and Physical Monitoring (Solaris ZFS)

Top  Previous  Next

This feature, along with the -MAP function, is part of the zfs aware version of smartmon-ux.  Basically, it automatically lights the appropriate LEDs based on the logical device status rather than physical device status.  For example, if a zpool is degraded, it will light the FAULT LED on the disk that is offline, and blink the rest of the drives in that pool if it is degraded.  All the disks in a pool that is being scrubbed will also blink during the scrub.

 

By default, the software will only use the Fault and identify LEDs, as they are found on the vast majority of SES enclosures, and tend to be consistently defined.  If you wish to take advantage of other LED states, then you should refer to the smartmon-ux.conf file which is used to cross-reference the vendor-specific LED codes, and the physical disk and pool status.

 

Invoke the automated monitoring and LED status as shown in the Usage section.  This will:

Automatically set the LEDs to reflect pool and disk status (-zfs)
Create the smartmon-ux.map file (which provides a cross-reference list for purposes of scripting)
Does a S.M.A.R.T. poll every 10 minutes / 600 seconds  (-F 600)
Reports status of enclosure health (-E)
Suppresses "OK" messages.  Only problems will be logged (-sq)

 

Usage:

smartmon-ux -zfs -MAP -F 600 -E -sq

 

Once you enter the above (or configure host O/S to enter the command at boot, then it is all automatic. The program will enumerate everything and relaunch in the background monitoring mode.  When the software polls the hardware and detects a state change, it will manipulate the LEDs accordingly and send out alerts.

Reported Pool States (In map file)

DEGRADED
FAULTED
OFFLINE
ONLINE
REMOVED
UNAVAIL
REMOVE
UNKNOWN

 

Reported Disk States (In map file)

FAULTED
OFFLINE
ONLINE
AVAIL
UNAVAIL
FOREIGN
INUSE
REMOVED