Methods and systems for computer monitoring
Methods and systems for deriving metrics for service and/or computer (e.g., server) utilization and stability from instrumentation of applications showing Poisson characteristics are disclosed. Such methods and systems may be further applied to automatically act upon the conditions indicated by the derived metrics, for example, to provision additional virtual server resource(s) in a cloud-based system.
1. A computer-based method for monitoring utilization of a server computer during a sample period, the server computer servicing received messages, the method comprising the steps of:
recording, in an electronic file, the time at which each message is dequeued and the time at which each message has been serviced;
calculating, using the computer, a mean service time and an arrival rate for the messages serviced during the sample period;
calculating, using the computer, a maximum throughput of the server based on the mean service time and a concurrency value;
calculating a server utilization based on the measured throughput and the calculated maximum throughput;
comparing the calculated server utilization to a utilization threshold to determine server overcapacity.
2. The method of claim 1 , further comprising the step of provisioning a server computer instance if the calculated server utilization exceeds the utilization threshold.
3. The method of claim 1 , wherein the concurrency value is stored in an electronic file of the server computer.
4. The method of claim 1 , wherein the concurrency value is a received parameter.
5. A computer-based method for monitoring the stability of a server computer, the method comprising the steps of:
recording, in an electronic file, the time at which each message is dequeued and the time at which each message has been serviced;
calculating, using the computer, a mean service time and a standard deviation of service times for the messages serviced during the sample period;
calculating a variability value based on the mean service time and stand deviation of service times;
calculating a stability value based on the variability value and a predetermined dampening factor; and
comparing the stability value to a stability threshold to determine server stability.