Abstract: Embodiments described herein are generally directed to testing an upgrade to a microservice in a production environment. According to an example, a first domain specific language (DSL) program is loaded into a kernel that wraps a container of a production environment of a container orchestration system hosting a known working version of the microservice. Production interactions with the microservice are intercepted by the first DSL program and state information associated with the production interactions is persisted. A second DSL program is loaded into a kernel that wraps a container of the production environment hosting the upgraded microservice. The upgraded microservice is isolated from the production environment by using the persisted state information as a source of data for all requests issued by the upgraded microservice. Outputs of the upgraded microservice and the known working version of the microservice are compared by replaying the production interactions to the upgraded microservice.

Abstract: Embodiments described herein are generally directed to testing an upgrade to a microservice in a production environment. According to an example, a first domain specific language (DSL) program is loaded into a kernel that wraps a container of a production environment of a container orchestration system hosting a known working version of the microservice. Production interactions with the microservice are intercepted by the first DSL program and state information associated with the production interactions is persisted. A second DSL program is loaded into a kernel that wraps a container of the production environment hosting the upgraded microservice. The upgraded microservice is isolated from the production environment by using the persisted state information as a source of data for all requests issued by the upgraded microservice. Outputs of the upgraded microservice and the known working version of the microservice are compared by replaying the production interactions to the upgraded microservice.

Abstract: Examples relate to managing file formats. In one example, a computing device may identify a particular file having a first format; generate a reference file for a second format of the particular file, the reference file including i) metadata for the second format of the particular file, and ii) at least one reference to a location of data included in the particular file; and generate, for the particular file, a file map that identifies the first format as a base file for the particular file.