NonStop Operations Guide for H-Series and J-Series RVUs

SLICE bladeId
is the identification of the Blade Element from which the processor element is to be dumped.
Valid values are A or B or C or ALL. Note that ALL may not be used with the parallel
method of dumping.
START n...
is the byte address where the dump will start. The default value is 0.
END n...
is the byte address where the dump will stop. Using a value of -1 is the same as specifying
the end of memory. The default value is -1.
ONLINE
If this option is specified, a dump can be taken of a processor while it is running. You
may use either PARALLEL or ONLINE but not both.
PARALLEL
If this option is specified, a dump may be taken of a single processor element while the
other PEs in that logical processor are reloaded and continue normal operations. You may
use either PARALLEL or ONLINE but not both.
For more information, see the HP NonStop TACL Reference Manual.
FULL
suppresses creation of an extracted dump file and results in creation of only a full dump
file.
For more information, see the HP NonStop TACL Reference Manual.
3. Monitor the dump to make sure that it finishes successfully:
a. Wait for this message to appear:
Dump completed.
b. Check the size of dumpfile to verify that the end-of-file pointer (EOF) is not equal to
zero.
> FUP INFO dumpfile
When a processor is dumped to disk, the RCVDUMP utility begins copying the dump in
a compressed format from the specified processor into a disk file called dumpfile. If
dumpfile does not exist, the RCVDUMP utility creates it. As the dump proceeds, the
status of the processor being dumped changes in the Processor Status dialog box to
indicate that a dump is in progress. When the dump is finished, the status of the selected
processor in the Processor Status dialog box changes to indicate the completion of the
dump.
A processor can be dumped to as many a two files: the full dump file and the extracted
dump file. The file name of the full dump file may end with F.
Dumping a Processor Using RCVDUMP for NonStop NS16000 Series and NS14000 Series Systems
• If your service provider determines that a processor halt is divergence-related, you might be
directed to dump the entire processor before reloading it. In this case, use the RCVDUMP
command, as follows:
◦ Use SLICE ALL parameter option.
◦ Do not specify ONLINE or PARALLEL parameters.
136 Processors and Components: Monitoring and Recovery