In case this is useful for anyone but me, I attach a slightly changed external/ssh.cts, which will call "reboot.cts.sh", to do a kexec instead of an actual reboot.
If you happen to live on some scsi controller, this likely shortens the reboot time by several minutes. Used so far only on debian, but should work on other distros just as well, possibly with minor tweaks to the location and name of initrd and kernel image. -- : Lars Ellenberg : LINBIT | Your Way to High Availability : DRBD/HA support and consulting http://www.linbit.com DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.
#!/bin/sh # # External STONITH module for ssh. # # Copyright (c) 2004 SUSE LINUX AG - Lars Marowsky-Bree <[email protected]> # # This program is free software; you can redistribute it and/or modify # it under the terms of version 2 of the GNU General Public License as # published by the Free Software Foundation. # # This program is distributed in the hope that it would be useful, but # WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. # # Further, this software is distributed without any warranty that it is # free of the rightful claim of any third person regarding infringement # or the like. Any license provided herein, whether implied or # otherwise, applies only to this software file. Patent licenses, if # any, provided herein do not apply to combinations of this program with # other software, or any other product whatsoever. # # You should have received a copy of the GNU General Public License # along with this program; if not, write the Free Software Foundation, # Inc., 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. # SSH_COMMAND="/usr/bin/ssh -q -x -o PasswordAuthentication=no -o StrictHostKeyChecking=no -n -l root" #SSH_COMMAND="/usr/bin/ssh -q -x -n -l root" # reboot.cts.sh may do anything you like, # as long as it reliably resets the node, if called with -nf. # As this is intended only for use by the Cluster Test Suite, # it is usefull to reduce the reboot time as much as possible, # so you may for example implement it using kexec. REBOOT_COMMAND="echo 'sleep 2; /sbin/reboot.cts.sh -nf' | SHELL=/bin/sh at now >/dev/null 2>&1" # Warning: If you select this poweroff command, it'll physically # power-off the machine, and quite a number of systems won't be remotely # revivable. # TODO: Probably should touch a file on the server instead to just # prevent heartbeat et al from being started after the reboot. # POWEROFF_COMMAND="echo 'sleep 2; /sbin/poweroff -nf' | SHELL=/bin/sh at now >/dev/null 2>&1" POWEROFF_COMMAND="echo 'sleep 2; /sbin/reboot -nf' | SHELL=/bin/sh at now >/dev/null 2>&1" # Rewrite the hostlist to accept "," as a delimeter for hostnames too. hostlist=`echo $hostlist | tr ',' ' '` is_host_up() { for j in 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 do if ping -w1 -c1 "$1" >/dev/null 2>&1 then sleep 1 else return 1 fi done return 0 } case $1 in gethosts) for h in $hostlist ; do echo $h done exit 0 ;; on) # Can't really be implemented because ssh cannot power on a system # when it is powered off. exit 1 ;; off) # Shouldn't really be implemented because if ssh cannot power on a # system, it shouldn't be allowed to power it off. exit 1 ;; reset) h_target=`echo $2 | tr A-Z a-z` for h in $hostlist do h=`echo $h | tr A-Z a-z` [ "$h" != "$h_target" ] && continue if case ${livedangerously} in [Yy]*) is_host_up $h;; *) true;; esac then $SSH_COMMAND "$2" "$REBOOT_COMMAND" # Good thing this is only for testing... if is_host_up $h then exit 1 else exit 0 fi else # well... Let's call it successful, after all this is only for testing... exit 0 fi done exit 1 ;; status) if [ -z "$hostlist" ] then exit 1 fi for h in $hostlist do if ping -w1 -c1 "$h" 2>&1 | grep "unknown host" then exit 1 fi done exit 0 ;; getconfignames) echo "hostlist" exit 0 ;; getinfo-devid) echo "ssh STONITH device" exit 0 ;; getinfo-devname) echo "ssh STONITH external device" exit 0 ;; getinfo-devdescr) echo "ssh-based Linux host reset" echo "Fine for testing, but not suitable for production!" echo "Only reboot action supported, no poweroff, and, surprisingly enough, no poweron." exit 0 ;; getinfo-devurl) echo "http://openssh.org"; exit 0 ;; getinfo-xml) cat << SSHXML <parameters> <parameter name="hostlist" unique="1" required="1"> <content type="string" /> <shortdesc lang="en"> Hostlist </shortdesc> <longdesc lang="en"> The list of hosts that the STONITH device controls </longdesc> </parameter> <parameter name="livedangerously" unique="0" required="0"> <content type="enum" /> <shortdesc lang="en"> Live Dangerously!! </shortdesc> <longdesc lang="en"> Set to "yes" if you want to risk your system's integrity. Of course, since this plugin isn't for production, using it in production at all is a bad idea. On the other hand, setting this parameter to yes makes it an even worse idea. Viva la Vida Loca! </longdesc> </parameter> </parameters> SSHXML exit 0 ;; *) exit 1 ;; esac
reboot.cts.sh
Description: Bourne shell script
_______________________________________________________ Linux-HA-Dev: [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev Home Page: http://linux-ha.org/
