System for monitoring elastic cloud-based computing systems as a service
View Patent ↗Provided is a computing-system monitor configured to monitor a plurality of computing-systems each having a plurality of monitored computing-instances. The computing-system monitor may include a plurality of collectors, each collector executed by one of a plurality of monitored computing-instances, wherein the plurality of monitored computing-instances each are part of one of a plurality of separately monitored computing systems, and wherein each of the collectors is operable to output metrics of a corresponding monitored computing-instance executing that collector. The computing-system monitor may also include an analytics platform, the analytics platform having a plurality of analytic computing-instances, the analytics platform being operable to receive metrics output by the plurality of collectors, calculate statistics with the analytic computing-instances based on the received metrics, and output the calculated statistics.
1. A tangible, non-transitory, machine-readable medium storing instructions that when executed by one or more computer systems effectuate operations comprising:
receiving, over a network, from a plurality of monitored computing instances, encrypted metric batches from collectors executed by the monitored computing instances;
decrypting the metric batches from the collectors;
associating the metric batches with one or more accounts, each metric batch being associated with one of the accounts;
parsing the decrypted metric batches from a serialized data format into hierarchical key-value pairs each associated with a respective one of the accounts;
adding the parsed hierarchical key-value pairs to an analysis queue of data to be analyzed;
retrieving the parsed hierarchical key-value pairs from the queue;
updating, with a processor, each of a plurality of statistics each describing values of metrics over a different temporal window based on the retrieved parsed hierarchical key-value pairs, the statistics being associated with the same account with which the corresponding metric batch is associated; and
receiving a request from a user device to view statistics of metrics of monitored computing instances, the request being associated with one of the accounts; and
sending the updated statistics associated with the account of the request to the user device.
2. The medium of claim 1 , wherein the operations comprise:
receiving another request from another user device to view statistics of metrics of monitored computing instances, the other request being associated with another one of the accounts;
sending the updated statistics associated with the other account of the other request to the other user device for display, the account and the other account being different accounts associated with different monitored computing instances.
3. The medium of claim 1 , wherein the operations comprise:
determining, with a process analyzing a shorter temporal window, that a duration of time of a longer temporal window has transpired and, in response, initiating an analysis of metric values of the longer temporal window.
4. The medium of claim 1 , wherein a plurality of statistics are calculated for each of the temporal windows, and wherein each of the statistics is compared against alarm criteria.
5. The medium of claim 1 , wherein the operations comprise:
automatically increasing a number of instances of monitoring computing devices in response to an increase in a rate of receiving the metric batches.
6. The medium of claim 1 , wherein the operations comprise:
validating a subscription associated with the account of the request before sending the updated statistics; and
billing the account.
7. The medium of claim 1 , wherein the metrics batches comprise metric values indicative of processor usage, memory usage, and network usage by a respective monitored computing instance.
8. The medium of claim 1 , wherein the operations comprise:
grouping the parsed hierarchical key-value pairs according to user-defined groups associated with the same account as the parsed hierarchical key-value pairs; and
calculating statistics of the groups.