Microservices anomaly detection
Presented herein are system and methods for countermeasures to address anomalies in microservices. A server having one or more processors coupled with memory may receive a first plurality of metrics from a defined set of microservices for a function. The server may apply the first plurality of metrics to an ensemble of anomaly detection models to generate a plurality of classifications. Each classification may indicate the first plurality of metrics as one of anomalous or normal from a respective model of the ensemble of anomaly detection models. The server may identify a majority of the plurality of classifications as corresponding to an anomaly event. The server may determine, responsive to identifying the majority, that at least one of the first plurality of metrics satisfies a criterion of a policy of a plurality of policies. The server may perform a countermeasure identified by the policy to address the anomaly event.
1 . A method, comprising:
receiving, by a server, a first plurality of metrics over a first time period from a defined set of microservices for a function;
applying, by the server, the first plurality of metrics to an ensemble of anomaly detection models to generate a plurality of classifications, each of the plurality of classifications indicating the first plurality of metrics as one of anomalous or normal from a respective model of the ensemble of anomaly detection models, wherein the ensemble of anomaly detection models is trained using a second plurality of metrics over a second time period;
identifying, by the server, a majority of the plurality of classifications as corresponding to an anomaly event in the defined set of microservices;
determining, by the server, responsive to identifying the majority of the plurality of classifications as corresponding to the anomaly event, that (i) a number of instances of the defined set of microservices affected by the anomaly event is greater than a first threshold number and (ii) a total number of instances of the defined set of microservices is less than a second threshold number in accordance with a criterion of a policy of a plurality of policies, each of the plurality of policies identifying a respective countermeasure to address the anomaly event; and
performing, by the server, a countermeasure identified by the policy to provide an alert message identifying the anomaly event to prompt an administrator to invoke restarting of the defined set of microservices for the function.
2 . The method of claim 1 , wherein determining further comprises determining that (i) a number of instances of the defined set of microservices affected by the anomaly event is less than a first threshold number and (ii) a total number of instances of the defined set of microservices is greater than or equal to a second threshold number, in accordance with a second criterion of a second policy, and
wherein performing further comprises performing a second countermeasure including a restart of the defined set of microservices without approval from an administrator.
3 . The method of claim 1 , wherein determining further comprises determining that (i) a number of instances of the defined set of microservices affected by the anomaly event is greater than a threshold number and (ii) a time elapsed since a restarting of the defined set of microservices is less than a threshold time, in accordance with a second criterion of a second policy, and
wherein performing further comprises performing a second countermeasure to provide an alert message identifying the anomaly event to prompt the administrator for examination.
4 . The method of claim 1 , wherein applying further comprises:
applying a natural language processing (NLP) algorithm to log data identifying a plurality of events associated with the defined set of microservices in carrying out the function; and
identifying, from applying the NLP algorithm to the log data, an exception in at least one of the plurality of events of an exception type associated with the anomaly event.
5 . The method of claim 4 , further comprising generating, by the server, for the plurality of classifications, a classification to identify the exception in the defined set of microservices as one of anomalous or normal based on the exception type.
6 . The method of claim 1 , further comprising receiving, by the server, via a user interface, a web document to include in a plurality of web documents maintained on a database, each of the plurality of web documents identified using at least one of (i) an anomaly type, (ii) one or more microservices, or (iii) a function type.
7 . The method of claim 1 , further comprising selecting, by the server, responsive to identifying the majority of the plurality of classifications as corresponding to the anomaly event, a web article from a plurality of web documents on a database based at least one of: (i) an anomaly type of the anomaly event, (ii) the defined set of microservices, or (iii) the function; and
wherein performing further comprises performing the countermeasure to provide an identification of the web document.
8 . The method of claim 1 , further comprising:
determining, by the server, responsive to identifying the majority of a second plurality of classifications as corresponding to a second anomaly event, that none of a third plurality of metrics satisfies any criterion of any policy of the plurality of policies; and
refraining, by the server, from performing any countermeasure identified by any policy of the plurality of policies, responsive to determining that none of a third plurality of metrics satisfies any criterion.
9 . The method of claim 1 , wherein the defined set of microservices further comprises a group of microservices invoked in response to carry out a function.
10 . A system, comprising:
a server having one or more processors coupled with memory, configured to
receive a first plurality of metrics over a first time period from a defined set of microservices for a function;
apply the first plurality of metrics to an ensemble of anomaly detection models to generate a plurality of classifications, each of the plurality of classifications indicating the first plurality of metrics as one of anomalous or normal from a respective model of the ensemble of anomaly detection models, wherein the ensemble of anomaly detection models is trained using a second plurality of metrics over a second time period;
identify a majority of the plurality of classifications as corresponding to an anomaly event in the defined set of microservices;
determine, responsive to identifying the majority of the plurality of classifications as corresponding to the anomaly event, that (i) a number of instances of the defined set of microservices affected by the anomaly event is greater than a first threshold number and (ii) a total number of instances of the defined set of microservices is less than a second threshold number in accordance with a criterion of a policy of a plurality of policies, each of the plurality of policies identifying a respective countermeasure to address the anomaly event; and
perform a countermeasure identified by the policy to provide an alert message identifying the anomaly event to prompt an administrator to invoke restarting of the defined set of microservices for the function.
11 . The system of claim 10 , wherein the server is further configured to:
determine that (i) a number of instances of the defined set of microservices affected by the anomaly event is less than a first threshold number and (ii) a total number of instances of the defined set of microservices is greater than or equal to a second threshold number, in accordance with a second criterion of a second policy, and
perform a second countermeasure including a restart of the defined set of microservices without approval from an administrator.
12 . The system of claim 10 , wherein the server is further configured to:
determine that (i) a number of instances of the defined set of microservices affected by the anomaly event is greater than a threshold number and (ii) a time elapsed since a restarting of the defined set of microservices is less than a threshold time, in accordance with a second criterion of a second policy, and
perform a second countermeasure to provide an alert message identifying the anomaly event to prompt the administrator for examination.
13 . The system of claim 10 , wherein the server is further configured to:
apply a natural language processing (NLP) algorithm to log data identifying a plurality of events associated with the defined set of microservices in carrying out the function; and
identify, from applying the NLP algorithm to the log data, an exception in at least one of the plurality of events of an exception type associated with the anomaly event.
14 . The system of claim 13 , wherein the server is further configured to generate, for the plurality of classifications, a classification to identify the exception in the defined set of microservices as one of anomalous or normal based on the exception type.
15 . The system of claim 10 , wherein the server is further configured to receive, via a user interface, a web document to include in a plurality of web documents maintained on a database, each of the plurality of web documents identified using at least one of (i) an anomaly type, (ii) one or more microservices, or (iii) a function type.
16 . The system of claim 10 , wherein the server is further configured to:
select, responsive to identifying the majority of the plurality of classifications as corresponding to the anomaly event, a web article from a plurality of web documents on a database based at least one of: (i) an anomaly type of the anomaly event, (ii) the defined set of microservices, or (iii) the function; and
perform the countermeasure to provide an identification of the web document.
17 . The system of claim 10 , wherein the server is further configured to:
determine, responsive to identifying the majority of a second plurality of classifications as corresponding to a second anomaly event, that none of a third plurality of metrics satisfies any criterion of any policy of the plurality of policies; and
refrain from performing any countermeasure identified by any policy of the plurality of policies, responsive to determining that none of a third plurality of metrics satisfies any criterion.
18 . The system of claim 10 , wherein the defined set of microservices further comprises a group of microservices invoked in response to carry out a function.