摘要 |
<p>Sampling sensor data in a sensor network for monitoring a computing system, where primary sensor nodes operable to measure a primary parameter of one or more component(s): and secondary sensor node(s) operable to obtain secondary parameter. Based upon one or more determined or pre-determined relationships between secondary parameters and primary parameter, a sample rate is determined for said primary parameter, or for the second parameter where that can be used to directly infer the properties of the primary parameter. This may be used to reduce impact on processing, by using hard drive enclosure, mid-plane or CPU temperature, vibration, fan speed, I/O rate, power, hard drive lifetime, network bandwidth, customer usage, component versions, loading, memory or network usage, altitude, shock, particulates as an indicator of hard drive status which is monitored using SMART diagnostic data. A Bayesian probability analysis may be used.</p> |