Fujitsu M10 Product Notes page 141

Table 3-16 Problems resolved in XCP 2080 (continued)
SPARC M10-
RTI No. 1 4 4S
RTIF2- x x x
131023-001
RTIF2- x x x
131023-004
RTIF2- x x x
131108-002
RTIF2- x x x
131112-009
Description
If the XSCF is reset by the rebootxscf(8)
or switchscf(8) command executed while
the logical domain is either starting or
stopped, the following problems may
occur.
- If the showdomainstatus(8) command
is executed, the logical domain state is
not displayed properly.
- If the showlogs event command is
executed, the log used for notification
of the logical domain state is not
displayed.
- An SNMP Trap for notification of the
state of the logical domain is not sent.
Even when an I/O failure is detected in
the logical domain, the XSCF may not be
notified of an error. For this reason, the
I/O failure information in Fault Report to
be displayed from the execution of the
fmdump(1M) command is not displayed,
even if the showlogs error command is
executed.
Moreover, even if a CPU or memory
error is detected in the XSCF, no
notification may be sent to the logical
domain. For this reason, Fault Report
does not display the CPU or memory
failure information to be displayed from
the execution of the showlogs error
command, even if the fmdump(1M)
command is executed.
The warning message "WARNING:
invalid vector intr: number 0x10001, pil
0x11000" may be output to /var/adm/
messages.
If the physical partition (PPAR) is
powered on, the event log of "LDC
Protocol info from PPAR (PPAR ID
0:Domain Service Data Send Failed)" is
output. Moreover, after Oracle Solaris
activation, "ldoms/ldmd:default failed:
transitioned to maintenance" is output to
the OS console. This may cause Oracle
VM Server for SPARC to enter
maintenance mode.
Workaround
There is no effective workaround.
[How to restore]
To display the state of the logical domain
properly, update it such as by reactivating
the logical domain.
There is no effective workaround.
If an I/O failure is detected in the logical
domain, locate the failed portion from
the display in Fault Report by executing
the fmdump(1M) command, and perform
maintenance on the failed part. If a CPU
or memory failure is detected in the
XSCF, perform maintenance according to
the FRU displayed by executing the
showlogs error command.
Ignore this message.
There is no effective workaround.
[How to restore]
Execute the rebootxscf -a command to
reset all the XSCFs, and then reactivate
the control domain. If restoration fails,
execute the poweroff(8) and poweron(8)
commands on the XSCF to power off and
then power on the physical partitions
(PPARs).
Chapter 3 Information on Software 127