If savecore did not successfully run to retrieve system core image(crash dump) from the dump device during reboot following a system panic it is possible to attempt to retrieve the crash dump manually from the command line. The kernel core image gets copied into the dump device configured by the dumpadm command as part of the panic recovery code. The default dump device is primary swap space. Often, the kernel core image may remain intact for some time after the panic.
First, ensure there is sufficient space to save the core file. A kernel core file will not be larger than the amount of physical memory installed in the system, but most core dumps will be considerably smaller. Insufficient space is usually the reason the data on the dump device does not get copied to the savecore directory. The error/warning you may see in the messages file will be similar to
savecore: [ID 976488 auth.error] not enough space in /var/crash/rascal (49144 MB avail, 51525 MB needed)
The default destination for the saved core file is the directory configured by dumpadm. If an alternate destination is required then one can be given in the savecore command, see below. As root, run savecore from the command line, specifying an alternate destination directory if required.
# savecore -dv [new dump directory]