We have a VMWare host server running Red Hat Enterprise Linux 5 x86_64 on a fairly powerful machine. It has two virtual machines, one Windows Server 2003 and one 2008. Periodically the host server gets very bogged down and the load spikes to around 20, and the server becomes essentially unresponsive.
The server has two Intel Xeon E5405 CPUs running at 2GHz a piece and 4GB RAM.
9567 root 5 -10 1997m 74m 41m S 13 1.9 3630:38 vmware-vmx
9670 root 6 -10 2007m 67m 30m D 9 1.7 8162:27 vmware-vmx
7555 mysql 15 0 365m 15m 2960 S 7 0.4 28:53.68 mysqld
The interesting thing there is, neither the CPU nor RAM usage seem all that high. Is there anything I can do to regulate the peak resource usage of the vmware virtual machines, or information I can provide that would be diagnostically useful?