Automated creation of resiliency experiments
A system automatically generates resiliency tests to detect the resiliency of an application implemented over multiple machines. In operation, one or more services operating on a plurality of machines is automatically identified. Fault targets are then automatically created based on the identified one or more services. The generation of the fault targets may be based on the identified services, user input, or past performance of the plurality of machines. Experiments may then be conducted, based on the created targets, on the one or more services. The experiments may insert faults into resources associated with the services. The resources may include CPU resources, memory, networking resources, and APIs. A resiliency score may be generated based on the experiments.
1. A method for automatically generating resiliency tests, comprising:
automatically identifying one or more services operating on a plurality of machines, wherein identifying includes:
installing agents at the plurality of the machines,
intercepting, by each agent, traffic transmitted to each of the plurality of machines and from each of the one plurality of machines with respect to one or more resources,
generating traffic data that includes one or more resources located remote from the plurality of machines and engaged in traffic with one or more of the plurality of machines,
transmitting generated traffic data to an application by each agent, and
identifying the one or more services by the application based on the traffic data received from the agents;
automatically creating fault targets based on the identified one or more services;
automatically conducting experiments on the one or more services using the created fault targets; and
generating a resilience score based on the experiments.
2. The method of claim 1 , wherein creating fault targets includes:
identifying, for each identified service, one or more resources or network connections; and
creating a fault for each resource and network connection.
3. The method of claim 1 , wherein conducting experiments includes creating an experiment to insert a fault for one or more resources and one or more network connection associated with each of the one or more services.
4. The method of claim 1 , wherein conducting experiments includes:
identifying, for each identified service, one or more resources or network connections; and
receiving input from a user regarding faults to insert in the one or more services; and
creating an experiment based on the user input, wherein each experiment inserts two faults in one or more of the one or more services, the two faults selected based on the received user input.
5. The method of claim 4 , wherein the user input includes one or more of connections, workload, fault type, and connection type.
6. The method of claim 1 , wherein conducting experiments includes:
identifying, for each identified service, one or more resources or network connections;
retrieving a template that inserts a plurality of faults into a resource or network connection; and
creating an experiment based on retrieve template, wherein each experiment inserts a plurality of faults in one or more of the one or more services, the plurality of faults selected based on a received user input.
7. The method of claim 1 , further comprising automatically recommending one or more experiments for execution after creation.
8. The method of claim 7 , wherein the one or more experiments are recommended based at least in part on traffic observed for a service of the one or more services or the number of services connected to the service of the one or more services.
9. A non-transitory computer readable storage medium having embodied thereon a program, the program being executable by a processor to perform a method for automatically generating resiliency tests, the method comprising:
automatically identifying one or more services operating on a plurality of machines, wherein identifying includes:
installing agents at the plurality of the machines,
intercepting, by each agent, traffic transmitted to each of the plurality of machines and from each of the one plurality of machines with respect to one or more resources,
generating traffic data that includes one or more resources located remote from the plurality of machines and engaged in traffic with one or more of the plurality of machines,
transmitting generated traffic data to an application by each agent, and
identifying the one or more services by the application based on the traffic data received from the agents;
automatically creating fault targets based on the identified one or more services;
automatically conducting experiments on the one or more services using the created fault targets; and
generating a resilience score based on the experiments.
10. The non-transitory computer readable storage medium of claim 9 , wherein creating fault targets includes:
identifying, for each identified service, one or more resources or network connections; and
creating a fault for each resource and network connection.
11. The non-transitory computer readable storage medium of claim 9 , wherein conducting experiments includes creating an experiment to insert a fault for one or more resources and one or more network connection associated with each of the one or more services.
12. The non-transitory computer readable storage medium of claim 9 , wherein conducting experiments includes:
identifying, for each identified service, one or more resources or network connections; and
receiving input from a user regarding faults to insert in the one or more services; and
creating an experiment based on the user input, wherein each experiment inserts two faults in one or more of the one or more services, the two faults selected based on the received user input.
13. The non-transitory computer readable storage medium of claim 12 , wherein the user input includes one or more of connections, workload, fault type, and connection type.
14. The non-transitory computer readable storage medium of claim 9 , wherein conducting experiments includes:
identifying, for each identified service, one or more resources or network connections;
retrieving a template that inserts a plurality of faults into a resource or network connection; and
creating an experiment based on retrieve template, wherein each experiment inserts a plurality of faults in one or more of the one or more services, the plurality of faults selected based on a received user input.
15. A system for automatically generating resiliency tests, comprising:
one or more servers, wherein each server includes a memory and a processor; and
one or more modules stored in the memory and executed by at least one of the one or more processors to automatically identifying one or more services operating on a plurality of machines, wherein identifying includes, installing agents at the plurality of the machines, intercepting, by each agent, traffic transmitted to each of the plurality of machines and from each of the one plurality of machines with respect to one or more resources, generating traffic data that includes one or more resources located remote from the plurality of machines and engaged in traffic with one or more of the plurality of machines, transmitting generated traffic data to an application by each agent, and identifying the one or more services by the application based on the traffic data received from the agents, automatically creating fault targets based on the identified one or more services, automatically conducting experiments on the one or more services using the created fault targets, and generating a resilience score based on the experiments.
16. The system of claim 15 , wherein creating fault targets includes identifying, for each identified service, one or more resources or network connections, and creating a fault for each resource and network connection.
17. The system of claim 15 , wherein conducting experiments includes creating an experiment to insert a fault for one or more resources and one or more network connection associated with each of the one or more services.
18. The system of claim 15 , wherein conducting experiments includes identifying, for each identified service, one or more resources or network connections; and receiving input from a user regarding faults to insert in the one or more services, and creating an experiment based on the user input, wherein each experiment inserts two faults in one or more of the one or more services, the two faults selected based on the received user input.
19. The system of claim 15 , wherein conducting experiments includes identifying, for each identified service, one or more resources or network connections, retrieving a template that inserts a plurality of faults into a resource or network connection, and creating an experiment based on retrieve template, wherein each experiment inserts a plurality of faults in one or more of the one or more services, the plurality of faults selected based on a received user input.