Psrinfo Showing Faulty Vcpu after MB replacement.

Asset ID:	1-72-1635008.1
Update Date:	2017-10-05
Keywords:

Solution Type Problem Resolution Sure

Solution 1635008.1 : Psrinfo Showing Faulty Vcpu after MB replacement.

Applies to:

Sun SPARC Enterprise T5240 Server - Version All Versions to All Versions [Release All Releases]
Information in this document applies to any platform.
After replacing Motherboard, several vcpus remains as faulted.

Symptoms

psrinfo reported several faulted vcpus.

# psrinfo

72 faulted since 02/25/2014 08:57:15
73 faulted since 02/25/2014 08:57:15
74 faulted since 02/25/2014 08:57:15
75 faulted since 02/25/2014 08:57:15
76 faulted since 02/25/2014 08:57:15
77 faulted since 02/25/2014 08:57:15
78 faulted since 02/25/2014 08:57:15
79 faulted since 02/25/2014 08:57:15

# prtdiag -v Also shows the same threads as faulted.

1 1415 MHz SUNW,UltraSPARC-T2+ on-line

72 1415 MHz SUNW,UltraSPARC-T2+ faulted <<<<<-------------------
73 1415 MHz SUNW,UltraSPARC-T2+ faulted <<<<<-------------------
74 1415 MHz SUNW,UltraSPARC-T2+ faulted <<<<<-------------------
75 1415 MHz SUNW,UltraSPARC-T2+ faulted <<<<<-------------------
76 1415 MHz SUNW,UltraSPARC-T2+ faulted <<<<<-------------------
77 1415 MHz SUNW,UltraSPARC-T2+ faulted <<<<<-------------------
78 1415 MHz SUNW,UltraSPARC-T2+ faulted <<<<<-------------------
79 1415 MHz SUNW,UltraSPARC-T2+ faulted <<<<<-------------------
80 1415 MHz SUNW,UltraSPARC-T2+ on-line
81 1415 MHz SUNW,UltraSPARC-T2+ on-line

Changes

The MotherBoard was replaced because it got a faulted CPU with caused by ereports fault.cpu.ultraSPARC-T2plus.dcache with message SUN4V-8001-FP

Cause

These vcpus got faulted because the number of D-cache errors associated with them have exceeded acceptable levels.

The fault manager will attempt to remove all threads in the same core.

Solution

Run solaris command psradm to re-enable all vcpus

psradm -n -F .. for all faulted vcpus.

Attachments

This solution has no attachment