Rollback for dependency services in cloud native environment
Managing multiple chain transaction rollback on a native cloud environment in a Kubernetes distributed system which extends API capabilities to build compensate API mappings which are used to reverse the invocation chain and stores the mappings in a key value store. An embodiment of the present invention extracts the reverse invocation chain from a framework of libraries and mapping API to check whether rollback is necessary when a service fails. An embodiment of the present invention executes an entire whole rollback.
1. A method of managing multiple chain transaction rollback in a distributed system of at least one federation of a plurality of interconnected pods in communication with a master application programming interface gateway, the method comprising the steps of:
receiving, by a validator component of the master application programming interface gateway, an indication of a request of a service container of a pod of the plurality of interconnected pods as failed from an application programming interface gateway of the pod;
validating, by the validator component, failure of the service container of the pod by:
obtaining at least one failed transaction from a request chain of the request of the service container; and
reading application programming interface trace mappings from service containers within the request chain;
determining, by the validator component, whether all application programming interfaces of the service containers in the request chain have compensate transactions;
sending, by the validator component, the trace mappings of the service containers in the request chain that have compensate transactions, to a runner component of the master application programming interface gateway;
receiving, by the runner component of the master application programming interface gateway, trace mappings from the validator component;
retrieving, by the runner component, a service container needed to rollback by a timestamp sequence;
executing, by the runner component, the compensate transactions associated with the service containers in a rollback chain in the timestamp sequence to build the rollback chain; and
executing, by the runner component, the rollback chain.
2. The method of claim 1 , further comprising a distributed tracing server in communication with the runner component of the master application programming interface gateway and with at least a repository of rollback statuses of service containers and request chains.
3. The method of claim 1 , further comprising a repository of resources required for rollback of the rollback chain.
4. A computer program product for managing multiple chain transaction rollback comprising:
one or more computer-readable storage media having computer-readable program instructions stored on the one or more computer-readable storage media, said program instructions executable by a master application programming interface gateway computer in a distributed system of at least one federation of a plurality of interconnected pods in communication with the master application programming interface gateway computer, the master application programming interface gateway computer comprising at least one processor, one or more memories, a validator component, and a runner component, the program instructions executing a computer-implemented method comprising:
receiving, by the validator component of the master application programming interface gateway computer, an indication of a request of a service container of a pod of the plurality of interconnected pods as failed from an application programming interface gateway of the pod;
validating, by the validator component, failure of the service container of the pod by:
obtaining at least one failed transaction from a request chain of the request of the service container; and
reading application programming interface trace mappings from service containers within the request chain;
determining, by the validator component, whether all application programming interfaces of the service containers in the request chain have compensate transactions;
sending, by the validator component, the trace mappings of the service containers in the request chain that have compensate transactions, to a runner component of the master application programming interface gateway computer;
receiving, by the runner component of the master application programming interface gateway computer, trace mappings from the validator component;
retrieving, by the runner component, a service container needed to rollback by a timestamp sequence;
executing, by the runner component, the compensate transactions associated with the service containers in a rollback chain in the timestamp sequence to build the rollback chain; and
executing, by the runner component, the rollback chain.
5. The computer program product of claim 4 , further comprising a distributed tracing server in communication with the runner component of the master application programming interface gateway computer and with at least a repository of rollback statuses of service containers and request chains.
6. The computer program product of claim 4 , further comprising a repository of resources required for rollback of the rollback chain.
7. A computer system comprising:
at least one processor;
one or more memories;
a runner component;
a validator component;
one or more computer readable storage media having program instructions executing a computer-implemented method for managing multiple chain transaction rollback in a distributed system of at least one federation of a plurality of interconnected pods, each pod having at least a service container in communication with the computer system, the program instructions comprising:
receiving, by the validator component, an indication of a request of a service container of a pod of the plurality of interconnected pods as failed from an application programming interface gateway of the pod;
validating, by the validator component, failure of the service container of the pod by:
obtaining at least one failed transaction from a request chain of the request of the service container; and
reading application programming interface trace mappings from service containers within the request chain;
determining, by the validator component, whether all application programming interfaces of the service containers in the request chain have compensate transactions;
sending, by the validator component, the trace mappings of the service containers in the request chain that have compensate transactions, to the runner component;
receiving, by the runner component, trace mappings from the validator component;
retrieving, by the runner component, a service container needed to rollback by a timestamp sequence;
executing, by the runner component, the compensate transactions associated with the service containers in a rollback chain in the timestamp sequence to build the rollback chain; and
executing, by the runner component, the rollback chain.
8. The computer system of claim 7 , further comprising a distributed tracing server in communication with the runner component, the distributing tracing server in communication with at least a repository of rollback statuses of service containers and request chains.
9. The computer system of claim 8 , wherein the distributed system is a Kubernetes distributed system.
10. The computer system of claim 8 , wherein the distributed system is in a cloud environment.
11. The computer system of claim 7 , further comprising a repository of resources required for rollback of the rollback chain.