Kalman filtering for grid computing telemetry and workload management
In one embodiment, a method and apparatus for Kalman filtering for grid computing telemetry and workload management is disclosed. The method includes monitoring telemetry data at a node of a grid computing engine with a Kalman filter, determining whether the monitored telemetry data is outside of a bounds of a predictive model of the Kalman filter, signaling an error condition for the node if the monitored telemetry data is outside of the bounds, and addressing the error condition by correcting a problem experienced at the node corresponding to the error condition. Other embodiments are also disclosed.
1. A method comprising:
monitoring telemetry data at a node of a grid computing engine with a Kalman filter;
determining whether the monitored telemetry data is outside of a bounds of a predictive model of the Kalman filter;
signaling an error condition for the node if the monitored telemetry data is outside of the bounds; and
addressing the error condition by correcting a problem experienced at the node corresponding to the error condition.
2. The method of claim 1 , wherein the Kalman filter recursively adjusts the predictive model based on the monitored telemetry data.
3. The method of claim 1 , wherein the determining whether the monitored telemetry data is outside of the bounds is performed in a real-time manner by the Kalman filter.
4. The method of claim 1 , wherein the Kalman filter does not save the monitored telemetry data.
5. The method of claim 1 , wherein the telemetry data includes at least one of a disk space of the node, a temperature of the node, a central processing unit (CPU) usage of the node, a transaction load of the node, and a network input/output (I/O) of the node.
6. The method of claim 5 , wherein more than one Kalman filter each monitors a specific type of the telemetry data at the node.
7. The method of claim 6 , further comprising utilizing a monitoring console to depict states of each of the more than one Kalman filter.
8. The method of claim 1 , wherein the Kalman filter is located at the node.
9. The method of claim 1 , wherein the Kalman filter is located at a server.
10. An article of manufacture, comprising a machine-readable storage medium including data that, when accessed by a machine, cause the machine to perform operations comprising:
monitoring telemetry data at a node of a grid computing engine with a Kalman filter;
determining whether the monitored telemetry data is outside of a bounds of a predictive model of the Kalman filter;
signaling an error condition for the node if the monitored telemetry data is outside of the bounds; and
addressing the error condition by correcting a problem experienced at the node corresponding to the error condition.
11. The article of manufacture of claim 10 , wherein the Kalman filter recursively adjusts the predictive model based on the monitored telemetry data.
12. The article of manufacture of claim 10 , wherein the determining whether the monitored telemetry data is outside of the bounds is performed by the Kalman filter in a real-time manner.
13. The article of manufacture of claim 10 , wherein the telemetry data includes at least one of a disk space of the node, a temperature of the node, a central processing unit (CPU) usage of the node, a transaction load of the node, and a network input/output (I/O) of the node.
14. The article of manufacture of claim 10 , wherein the Kalman filter is located at the node.
15. The article of manufacture of claim 10 , wherein the Kalman filter is located at a server.
16. An apparatus, comprising:
a grid computing engine having a node; and
a Kalman filter associated with the node to:
monitor telemetry data at the node;
determine whether the monitored telemetry data is outside of the bounds of a predictive model of the Kalman filter; and
signal an error condition for the node if the telemetry data is outside of the bounds.
17. The apparatus of claim 16 , further comprising an automated response system to address the error condition by correcting a problem experienced at the node corresponding to the error condition.
18. The apparatus of claim 16 , wherein the telemetry data includes at least one of a disk space of the node, a temperature of the node, a central processing unit (CPU) usage of the node, a transaction load of the node, and a network input/output (I/O) of the node.
19. The apparatus of claim 18 , wherein the Kalman filter recursively adjusts the predictive model based on the monitored telemetry data.
20. The apparatus of claim 16 , further comprising a monitoring console to depict a state of the Kalman filter.