MONITORING DAEMONSection: Rutgers DarkLab Tools (1)Updated: 2006-08-23 Index Return to Main Contents NAMEmonitord - temperature server emulator clientSYNOPSISmonitor [-s servername] [-p port] [-D delay] [-e script.energy] [-d] [-v] [-h]DESCRIPTIONMonitoring daemon. The monitor daemon, called monitord, periodically samples the utilization of the components of the machine on which it is running and reports that information to the solver. The components considered are the CPU(s), disk(s), and network interface(s) and their utilization information is computed from /proc. The frequency of utilization updates sent to the solver is a tunable parameter set to 1 second by default. Our current implementation uses 128-byte UDP messages to update the solver. Computing the amount of heat produced by a CPU using high-level utilization information may not be adequate for modern processors. For this reason, we have recently developed and validated a version of Mercury for the Intel Pentium 4 processor. In this version, monitord monitors the hardware performance counters and translates each observed performance event into an estimated energy. (The software infrastructure for handling the Intel performance counters has been provided by Frank Bellosa's group.) To avoid further modifications to Mercury, the daemon transforms the estimated energy during each interval into an average power, which is then linearly translated into a "low-level utilization" in the [0% = Pbase , 100% = Pmax ] range. This is the utilization reported to the solver. If Mercury is running online (not using traces), monitord should be running on every machine being emulated. One instance per machine per emulator server must be running for the emulator to give reliable results.
SEE ALSOemulator(1); fiddle(1); sensor(1);
IndexThis document was created by man2html using the manual pages. Time: 03:06:22 GMT, September 04, 2006 |