Tested and verified on 2.6.18-128.5.1.el5.
Test steps:
1. configure cluster w/o fencing methods
2. service cman start
3. service clvmd start
4. from nodeA, cman_tool kill -n nodeB
5. on nodeB run lvs
6. check top on nodeB for dlm_send using 99% of CPU
7. check slabtop on nodeB for TCP slab growth (it should not)
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.
http://rhn.redhat.com/errata/RHSA-2009-1222.html