[Pacemaker] 'crm configure verify' says my node(s) don't exist, everything else says they do

[mark - pacemaker list]

This output kind of shows it all... I can configure the cluster, put nodes
in standby and back online, move resources from one to the other, etc., but
if any rule references a node in the configuration, then 'crm configure
verify' fails saying the node doesn't exist.  This is a freshly-started
cluster, and both nodes have entries in /etc/hosts giving their shortname
and fqdn (shortname first, if it matters):

[root at kvm2 ~]# crm
crm(live)# configure show
node $id="1023518912" kvm1
node $id="1040296128" kvm2 \
        attributes standby="off"
primitive winsrv1 ocf:heartbeat:VirtualDomain \
        params config="/etc/kvm-configs/winsrv1.xml" \
        meta allow-migrate="true" \
        op start interval="0" timeout="90s" \
        op stop interval="0" timeout="180s" \
        op monitor interval="15s" timeout="30s" \
        op migrate_to interval="0" timeout="240s" \
        op migrate_from interval="0" timeout="240s" \
        utilization cpu="2" hv_memory="4096"
location cli-prefer-winsrv1 winsrv1 \
        rule $id="cli-prefer-rule-winsrv1" inf: #uname eq kvm1
property $id="cib-bootstrap-options" \
        dc-version="1.1.7-ee0730e13d124c3d58f00016c3376a1de5323cff" \
        cluster-infrastructure="corosync" \
        stonith-enabled="false" \
        last-lrm-refresh="1344975603"
crm(live)# configure verify
WARNING: cli-prefer-winsrv1: referenced node kvm1 does not exist
crm(live)# node show
kvm1(1023518912): normal
kvm2(1040296128): normal
        standby: off
crm(live)# node standby kvm2
crm(live)# status
============
Last updated: Tue Aug 14 15:29:26 2012
Last change: Tue Aug 14 15:27:26 2012 via crm_resource on kvm2
Stack: corosync
Current DC: kvm1 (1023518912) - partition with quorum
Version: 1.1.7-ee0730e13d124c3d58f00016c3376a1de5323cff
2 Nodes configured, unknown expected votes
1 Resources configured.
============

Node kvm2 (1040296128): standby
Online: [ kvm1 ]

 winsrv1        (ocf::heartbeat:VirtualDomain): Started kvm1
crm(live)# node online kvm2
crm(live)# resource unmove winsrv1
crm(live)# resource move winsrv1 kvm2
crm(live)# configure show
node $id="1023518912" kvm1
node $id="1040296128" kvm2 \
        attributes standby="off"
primitive winsrv1 ocf:heartbeat:VirtualDomain \
        params config="/etc/kvm-configs/winsrv1.xml" \
        meta allow-migrate="true" \
        op start interval="0" timeout="90s" \
        op stop interval="0" timeout="180s" \
        op monitor interval="15s" timeout="30s" \
        op migrate_to interval="0" timeout="240s" \
        op migrate_from interval="0" timeout="240s" \
        utilization cpu="2" hv_memory="4096"
location cli-prefer-winsrv1 winsrv1 \
        rule $id="cli-prefer-rule-winsrv1" inf: #uname eq kvm2
property $id="cib-bootstrap-options" \
        dc-version="1.1.7-ee0730e13d124c3d58f00016c3376a1de5323cff" \
        cluster-infrastructure="corosync" \
        stonith-enabled="false" \
        last-lrm-refresh="1344975603"
crm(live)# configure verify
WARNING: cli-prefer-winsrv1: referenced node kvm2 does not exist
crm(live)#


Did I miss something obvious?

Thanks for any pointers,
Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.clusterlabs.org/pipermail/pacemaker/attachments/20120814/b9b2a049/attachment-0002.html>