[Pacemaker] Remote node did not respond

[lepace]

Hi all,
I have set up a two-node cluster, at first, it works fine, but when I write a lot of data to the fs，a resource of pacemaker，when the filesystem is almost full, it collapse. Run crm_mon and it appears:
============
Last updated: Mon Jun 21 16:57:09 2010
Stack: openais
Current DC: mds2 - partition with quorum
Version: 1.0.5-462f1569a43740667daf7b0f6b521742e9eb8fa7
2 Nodes configured, 2 expected votes
6 Resources configured.
============


Online: [ mds1 mds2 ]


ipmi_mds1(stonith:external/ipmi) Started [mds1    mds2 ]
ipmi_mds2(stonith:external/ipmi):        Started mds1 FAILED
Resource Group: web_server
    virtual_ip  (ocf::heartbeat:IPaddr) Started [mds1    mds2 ]
    apache(ocf::heartbeat:apache) Started [mds1    mds2 ]
Clone Set: pingd_manage_net
    manage_pingd:0(ocf::pacemaker:pingd): Started mds1 FAILED
        Started: [ mds2 ]
Clone Set: pingd_data_net
    data_pingd:0        (ocf::pacemaker:pingd): Started mds1 FAILED
        Started: [ mds2 ]
metavol_mpath0  (ocf::heartbeat:Filesystem) Started [   mds1    mds2 ]


Failed actions:
    ipmi_mds1_monitor_0 (node=mds1, call=-1, rc=1, status=Timed Out): unknown error
    ipmi_mds2_monitor_0 (node=mds1, call=-1, rc=1, status=Timed Out): unknown error
    virtual_ip_monitor_0 (node=mds1, call=-1, rc=1, status=Timed Out): unknown error
    apache_monitor_0 (node=mds1, call=-1, rc=1, status=Timed Out): unknown error
    manage_pingd:0_monitor_0 (node=mds1, call=-1, rc=1, status=Timed Out): unknown error
    data_pingd:0_monitor_0 (node=mds1, call=-1, rc=1, status=Timed Out): unknown error
    metavol_mpath0_monitor_0 (node=mds1, call=-1, rc=1, status=Timed Out): unknown error


and then I run crm resource start ipmi_mds2 , and wait about 5 minutes, then it says :Error performing operation: Remote node did not respond


Which reason? And how can I avoid this situation?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20100621/9cae40db/attachment-0002.html>