savecore reads data from the disk dump partition specified by the
device argument, and
creates a vmcore file in a directory named /var/crash/127.0.0.1-
timestamp. The timestamp portion of the directory name indicates the date and time when
system crash occurred. The vmcore file can be analyzed by the
crash(8) utility.
savecore returns 1 if it fails due to an error. Otherwise it returns 0.
OPTIONS
-D
Delete dump from the dump partition after saved.
-f
Create a vmcore file forcibly even if dump directory exists.
-p
Display progress for creating a vmcore file.
-P flags
Specify the value of page flags. If the some flags of page have been matched
with the specified value, the page is saved in vmcore with uncompressed data.
This option is effective only when dump compression is used.
-v
Provide some verbose outputs.
-V
Show version information and exit.
FILES
/var/crash
The main directory where the crash dump files are stored. Each dump is put in
a subdirectory whose name consists of the concatenation of "127.0.0.1" and
the date and time of the crash.
/var/crash/scripts
This directory can contain scripts that are run at various times. They all get
passed the ip of 127.0.0.1 as the first argument, and each one gets
the directory that the dumpfile is written to as the second argument.
diskdump-nospace - This is run when there is not enough disk space for the dump
of the crashed machine. If this script exits with a non-zero return value,
savecore will try once again (but only once) before giving up on the dump. If this
script exits with a zero return value,
savecore will exit without performing a dump. The default script simply sends a mail
message and exits with a zero value, but it can be enhanced to try to free up
diskspace and return a non-zero value.
