table of contents
| DUMPCONF(8) | System Manager's Manual | DUMPCONF(8) |
NAME¶
dumpconf - Configure panic and PSW restart actions for Linux on System z
SYNOPSIS¶
dumpconf [start|stop|status]
dumpconf [-h|-v]
DESCRIPTION¶
dumpconf reads the /etc/sysconfig/dumpconf file and establishes the action to be taken if a kernel panic occurs or PSW restart is triggered.
The following keywords can be used in the dumpconf file:
- - ON_PANIC:
- Shutdown action in case of a kernel panic or a PSW restart. Possible
values are 'dump', 'reipl', 'dump_reipl', 'stop' and 'vmcmd':
dump: Trigger dump according to the configuration in /etc/sysconfig/dumpconf.
reipl: Trigger re-IPL according to the configuration under /sys/firmware/reipl.
dump_reipl: First trigger dump according to the configuration in /etc/sysconfig/dumpconf, then trigger re-IPL according to the configuration under /sys/firmware/reipl.
stop: Stop Linux and enter disabled wait (default).
vmcmd: Trigger CP command according to the 'VMCMD_X' configuration in /etc/sysconfig/dumpconf.
- - DUMP_TYPE:
- Type of dump device. Possible values are 'ccw', 'eckd', 'fcp' and 'nvme'.
- - DEVICE:
- Device number of dump device.
- - WWPN
- WWPN for SCSI dump device.
- - LUN
- LUN for SCSI dump device.
- - FID
- Function ID for NVMe dump device.
- - NSID
- Namespace ID for NVMe dump device.
- - BOOTPROG:
- Boot program selector.
- - BR_CHR:
- Boot record location in "C,H,R" format (comma separated values for Cylinder, Head and Record) or "auto".
- - BR_LBA:
- Boot record logical block address.
- - SCP_DATA:
- SCP data for SCSI, NVMe and ECKD dump devices.
- - VMCMD_1, VMCMD_2 ... VMCMD_8:
- Up to eight CP commands, which are executed in case of a kernel panic or PSW restart.
- - DELAY_MINUTES:
- Number of minutes the activation of dumpconf is to be delayed. If this
keyword is omitted, the default is zero, which means that dumpconf
activates immediately during system startup. Specify a non-zero delay time
only if you specified shutdown action "reipl" or
"dump_reipl". These actions might cause a reboot loop if the
Linux kernel crashes persistently during (or shortly after) each reboot.
A non-zero delay time causes dumpconf to sleep in the background until the delay time has expired. In this case messages are written to /var/log/messages. By default (DELAY_MINUTES is omitted or zero) dumpconf runs in the foreground and informational messages are written to sysout, while error messages are written to syserr.
Example: If you specified DELAY_MINUTES=10 and your Linux system crashes within 10 minutes after the reboot, then dumpconf is not yet active and the default action (stop) is triggered.
COMMANDS¶
OPTIONS¶
- -h or --help
- Print usage information, then exit.
- -v or --version
- Print version information, then exit.
PSW Restart¶
PSW Restart can be triggered by the operator under z/VM with the CP command "#cp system restart" and under LPAR on the HMC with "Recovery/PSW Restart".
EXAMPLES:¶
The following are examples of the /etc/sysconfig/dumpconf file:
#
# Example configuration for a CCW dump device (DASD)
#
ON_PANIC=dump_reipl
DUMP_TYPE=ccw
DEVICE=0.0.1234
DELAY_MINUTES=5
#
# Example configuration for an ECKD dump device (DASD)
#
ON_PANIC=dump
DUMP_TYPE=eckd
DEVICE=0.0.1004
BOOTPROG=0
BR_CHR=auto
#
# Example configuration for an FCP dump device (SCSI Disk)
#
ON_PANIC=dump
DUMP_TYPE=fcp
DEVICE=0.0.2345
WWPN=0x5005076303004712
LUN=0x4713000000000000
BOOTPROG=0
BR_LBA=0
#
# Example configuration for an NVMe dump device (NVMe Disk)
#
ON_PANIC=dump
DUMP_TYPE=nvme
FID=0x0300
NSID=0x0001
BOOTPROG=0
BR_LBA=0
#
# Example configuration for CP commands
#
ON_PANIC=vmcmd
VMCMD_1="MESSAGE * Starting VMDUMP"
VMCMD_2="VMDUMP"
VMCMD_3="IPL 3456"
#
# Example config for re-IPL
#
ON_PANIC=reipl
DELAY_MINUTES=5
SEE ALSO¶
Linux on System z: Using the Dump Tools
| Sept 2011 | s390-tools |