For these specific servers, the enumeration of internal storage can occur during a short window when Oracle Solaris is unable to successfully load new kernel driver modules.
If this occurs, Oracle Solaris will be unable to manage your internal storage. This includes the loss of fmd(1M)'s monitoring of disk SMART failures. The most observable symptom of this issue is having diskinfo(1M) output fail to show /SYS/HDD disks.
If the results of running this command show that you have a 'SPARC-S7-2L' system with either a 'SPARC-S7-2L-12dbp' or 'SPARC-S7-2L-24dbp' internal storage backplane, your server might be affected by this issue.
# /usr/sbin/devprop compatible
chassis,Sun-Microsystems.SPARC-S7-2L-24dbp.unknown.unknown +
chassis,Sun-Microsystems.SPARC-S7-2L-24dbp + chassis.SPARC-S7-2L-24dbp +
chassis,Sun-Microsystems.SPARC-S7-2L.unknown.unknown + chassis,Sun-Microsystems.SPARC-S7-2L +
chassis.SPARC-S7-2L + sun4v
# /usr/sbin/devprop compatible
chassis,Sun-Microsystems.SPARC-S7-2L-12dbp.unknown.unknown +
chassis,Sun-Microsystems.SPARC-S7-2L-12dbp + chassis.SPARC-S7-2L-12dbp +
chassis,Sun-Microsystems.SPARC-S7-2L.unknown.unknown + chassis,Sun-Microsystems.SPARC-S7-2L +
chassis.SPARC-S7-2L + sun4v
Workaround: The preinstalled OS (Oracle Solaris 11.3 SRU 9) contains a workaround for this issue, in the form of a line added to /etc/system: forceload: drv/ses.
This line is required for CR 18712182.
Do not remove until an SRU with a fix for 18712182 has been installed.
If you reinstall the server with Oracle Solaris 11.3 SRU 9, ensure that you add the corresponding line to /etc/system after re-installation. Once an SRU containing the fix for this CR 18712182 has been installed, remove this line from /etc/system.
Reboot the server to activate the changes to the /etc/system file.