User Tools

Site Tools


unix:freebsd:system_builds:carp

Differences

This shows you the differences between two versions of the page.


unix:freebsd:system_builds:carp [2019/07/10 20:46] (current) – created - external edit 127.0.0.1
Line 1: Line 1:
 +====== Setting up CARP ======
 +
 +You need carp set up to tell the nodes which is primary and which is secondary.
 +
 +I'm using vlan50 for my data, so we need to be able to switch between machines. The two machines will have 192.168.209.0/24 in common, configured as follows:
 +  * iscsi-0: 192.168.209.67
 +  * iscsi-1: 192.168.209.68
 +  * CARP: 192.168.209.64
 +
 +===== Enable carp and set shared IP =====
 +
 +Add the directives to load the carp module and make the alias for vlan50. the carp.preempt sets it so so when the parent interface goes down or comes up, the carp interface will do the same
 +<code bash>
 +echo 'carp_load="YES"' >> /boot/loader.conf
 +echo 'net.inet.carp.preempt=1' >> /etc/sysctl.conf
 +echo 'ifconfig_vlan50_alias0="inet vhid 1 pass iscsilan alias 10.19.209.64/24"' >> /etc/rc.conf
 +</code>
 +
 +if you want to test this, you can run the following:
 +
 +<code bash>
 +kldload carp  
 +sysctl net.inet.carp.preempt=1
 +ifconfig vlan50 alias vhid 1 pass iscsilan 10.19.209.64/24
 +</code>
 +
 +===== Set devd =====
 +
 +Add the following to the bottom of /etc/devd.conf. This tells the server what to do when the carp interface comes up or goes down.
 +<code bash>
 +notify 30 {
 +        match "system" "IFNET";
 +        match "subsystem" "vlan50";
 +        match "type" "LINK_UP";
 +        action "/usr/local/opt/hast/failover master";
 +};
 +
 +notify 30 {
 +        match "system" "IFNET";
 +        match "subsystem" "vlan50";
 +        match "type" "LINK_DOWN";
 +        action "/usr/local/opt/hast/failover slave";
 +};
 +
 +</code>
 +
 +===== Create the failover script =====
 +
 +This is taken mainly from the hast article at [https://opensource-information.blogspot.com/2012/08/hast-freebsd-zfs-with-carp-failover.html].
 +
 +The following script is designed to bring up/shutdown the hast volumes and will eventually be used to handle the iSCSI service and NFS. It already handles the zpool itself.
 +
 +Change resource list and pool name before implementing.
 +
 +<code bash failover>
 +#!/bin/sh
 +
 +# https://opensource-information.blogspot.com/2012/08/hast-freebsd-zfs-with-carp-failover.html
 +# Original script by Freddie Cash <fjwcash@gmail.com>
 +# Modified by Michael W. Lucas <mwlucas@BlackHelicopters.org>
 +# and Viktor Petersson <vpetersson@wireload.net>
 +# Modified by George Kontostanos <gkontos.mail@gmail.com>
 + 
 +# The names of the HAST resources, as listed in /etc/hast.conf
 +resources="disk0 disk1 disk2 disk3 disk4 disk5 disk6 disk7 disk8 disk9"
 + 
 +# delay in mounting HAST resource after becoming master
 +# make your best guess
 +delay=3
 + 
 +# logging
 +log="local0.debug"
 +name="failover"
 +pool="storage"
 + 
 +# end of user configurable stuff
 + 
 +case "$1" in
 + master)
 +  logger -p $log -t $name "Switching to primary provider for ${resources}."
 +  sleep ${delay}
 + 
 +  # Wait for any "hastd secondary" processes to stop
 +  for disk in ${resources}; do
 +   while $( pgrep -lf "hastd: ${disk} \(secondary\)" > /dev/null 2>&1 ); do
 +    sleep 1
 +   done
 + 
 +   # Switch role for each disk
 +   hastctl role primary ${disk}
 +   if [ $? -ne 0 ]; then
 +    logger -p $log -t $name "Unable to change role to primary for resource ${disk}."
 +    exit 1
 +   fi
 +  done
 + 
 +  # Wait for the /dev/hast/* devices to appear
 +  for disk in ${resources}; do
 +   for I in $( jot 60 ); do
 +    [ -c "/dev/hast/${disk}" ] && break
 +    sleep 0.5
 +   done
 + 
 +   if [ ! -c "/dev/hast/${disk}" ]; then
 +    logger -p $log -t $name "GEOM provider /dev/hast/${disk} did not appear."
 +    exit 1
 +   fi
 +  done
 + 
 +  logger -p $log -t $name "Role for HAST resources ${resources} switched to primary."
 + 
 + 
 +  logger -p $log -t $name "Importing Pool"
 +  # Import ZFS pool. Do it forcibly as it remembers hostid of
 +                # the other cluster node.
 +                out=`zpool import -f "${pool}" 2>&1`
 +                if [ $? -ne 0 ]; then
 +                    logger -p local0.error -t hast "ZFS pool import for resource ${resource} failed: ${out}."
 +                    exit 1
 +                fi
 +                logger -p local0.debug -t hast "ZFS pool for resource ${resource} imported."
 + 
 + ;;
 + 
 + slave)
 +  logger -p $log -t $name "Switching to secondary provider for ${resources}."
 + 
 +  # Switch roles for the HAST resources
 +  zpool list | egrep -q "^${pool} "
 +         if [ $? -eq 0 ]; then
 +                 # Forcibly export file pool.
 +                 out=`zpool export -f "${pool}" 2>&1`
 +                  if [ $? -ne 0 ]; then
 +                         logger -p local0.error -t hast "Unable to export pool for resource ${resource}: ${out}."
 +                         exit 1
 +                  fi
 +                 logger -p local0.debug -t hast "ZFS pool for resource ${resource} exported."
 +         fi
 +  for disk in ${resources}; do
 +   sleep $delay
 +   hastctl role secondary ${disk} 2>&1
 +   if [ $? -ne 0 ]; then
 +    logger -p $log -t $name "Unable to switch role to secondary for resource ${disk}."
 +    exit 1
 +   fi
 +   logger -p $log -t $name "Role switched to secondary for resource ${disk}."
 +  done
 + ;;
 +esac
 +</code>
 +===== References =====
 +  * https://opensource-information.blogspot.com/2012/08/hast-freebsd-zfs-with-carp-failover.html
 +  * https://wiki.freebsd.org/HAST
 +  * https://www.freebsd.org/doc/handbook/carp.html
  
unix/freebsd/system_builds/carp.txt · Last modified: 2019/07/10 20:46 by 127.0.0.1