[Rdo-list] Basic HA deployment

[Marius Cornea]

Hi everyone, 

I tried a deployment on virt with 3 x ctrls + 1 x compute and it currently fails due to a ceilometer dbsync issue(BZ#1271002). To workaround it I did the following. This gets the deployment successful but some of the Neutron related pacemaker resources are stopped(same as BZ#1270964):

1. Mount the overcloud-full.qcow2 image on a host with libguestfs-tools installed (I used the physical machine where I run the virt env for this)

guestfish --rw -a overcloud-full.qcow2
><fs> run
><fs> mount /dev/sda /
><fs> vi /etc/puppet/modules/ceilometer/manifests/init.pp
#Apply the changes below:
diff -c2 init.pp.orig init.pp.new 
*** init.pp.orig	2015-10-13 14:35:57.514488094 +0000
--- init.pp.new	2015-10-13 14:35:01.614488094 +0000
***************
*** 154,157 ****
--- 154,158 ----
    $qpid_reconnect_interval_max        = 0,
    $qpid_reconnect_interval            = 0,
+   $mongodb_replica_set                = 'tripleo', 
  ) {
  
***************
*** 293,296 ****
--- 294,298 ----
      'database/metering_time_to_live'      : value => $metering_time_to_live;
      'database/alarm_history_time_to_live' : value => $alarm_history_time_to_live;
+     'database/mongodb_replica_set'        : value => $mongodb_replica_set;
    }
><fs> quit

2. Get the overcloud-full.qcow2 image back on the undercloud and update the existing Glance image:
openstack overcloud image upload --update-existing

3. Deploy overcloud:
openstack overcloud deploy --templates ~/templates/my-overcloud -e ~/templates/my-overcloud/environments/network-isolation.yaml -e ~/templates/network-environment.yaml --control-scale 3 --compute-scale 1 --libvirt-type qemu -e /usr/share/openstack-tripleo-heat-templates/environments/puppet-pacemaker.yaml --ntp-server clock.redhat.com

Thanks,
Marius