More info on this. We are running CentOS 4.4 x86. Sorry I didn't include that before. We believe that this started happening about the time of the CentOS 4.4 kernel upgrade to 2.6.9-42.0.2.ELsmp. I'm writing info on the process using the most physical memory (rss) each minute to a log file. In the same minute as the error message, a user started a vmware workstation session. I'm thinking that this is a likely culprit. The rss number did not look unusual, though. About 98M RSS. Thanks, Steve _______________________________________________ CentOS mailing list CentOS@xxxxxxxxxx http://lists.centos.org/mailman/listinfo/centos