Fujitsu PRIMEQUEST 500A Series Installation Manual page 307

4.9.3.7 Checking kdump collection
If the system actually crashes, the dump may not be collected because of an area
shortage. To prevent this problem, check in advance whether the crash dump is
actually collected by using the following procedure:
Manually collect dump
Check for dump completion
Reboot system
Check dump collection
Delete dump
Note:
(1) Manually collecting a dump
Select [Partition] → [Power Control] from the MMB Web-UI to set the [INIT]
control, and then click the [Apply] button to start dump collection.
(2) Confirming that dump collection is complete
Confirm that dump collection has been completed with the following message
displayed on the console:
makedumpfile Completed.
Note:
The dumpfile is saved to /mnt/127.0.0.1-2007-02-14-15:20:37/vmcore-
incomplete.
C122-E001-10EN
If the crash dump is forcibly collected, the system is forced to crash. Before
verifying the dump collection, stop any important applications.
Unnecessary file systems should also be demounted.
When the dump has been collected successfully, the message shown in the
example that follows may be output. (Note that the actual dump file name is
not vmcore-incomplete but is vmcore.)
4.9 Dump Environment Setup (Linux: Red Hat)
4-121