If POST runs as part of the boot sequence and if the system experiences hardware faults, you might see a POST ERROR message, followed by an INFO message. The boot sequence continues, but the system might contain untested hardware. The untested hardware will be available to the operating system after the boot.
For these cases, POST stalls for several minutes, then sends an e-report, as in the following example:
2013-01-23 15:30:55:990 0:0:0>ERROR:
ereport.chassis.post.io.test-fail@/SYS/MB/CM0/CMP/PCIE_LINK1
reporting PCPU ID=0
TestTitle=IO Trap Handler
Operation=Data Access Error (Type: 32)
Trap PC=0x0000000000560b8c
Trap Level=0x01
NPESR=0x0000000000000001
NPEAR=0x0000805100700000
END_ERROR
2013-01-23 15:30:56:054 0:0:0>INFO: Link Down Recovery Not Supported
2013-01-23 15:31:02 0:0:0> ERROR: POST Timed out. Not all components tested
Workaround A: Use the fmdump -eV command to review the error logs and console messages. Replace or repair any failing hardware.
Workaround B: Restart POST. POST will be run on the system with the faulty components already deconfigured.