pablochacin
pablochacin
In order to facilitate the debugging of issues, provide a logging mechanism by means of the [Execution environment](#172)
When testing fault injection in the [Online Boutique](https://github.com/GoogleCloudPlatform/microservices-demo) application the requests seems to have a significant delay (around `500ms`) even when the faults did not specify any delay (only an...
Exposing multiple services in the clusters used for e2e testing has become increasingly complex as the number of tests increases and new protocols are introduced. The simplest way is to...
Cache servers are a fundamental component of modern applications. They are critical in providing low-latency access to information and preventing the overloading of databases. Providing fault injection capabilities to the...
Implement a disruptor for injecting faults in nodes of a Kubernetes cluster. The disruptor should allow selecting/excluding nodes based on labels, taints (e.g. `unschedulable`), and conditions (e.g. `Ready`). The disruptor...
MySQL is one of the most widely used database backends for modern applications. Providing fault injection capabilities for the MySQL protocol used between clients and servers will help developers to...
Kafka is widely used in microservices architecture to support async event-driven communication between services. Therefore, it is critical for teams to test how their services handle different types of errors...
Unit tests in the `disruptor_test.go` file are mostly redundant with the tests in the `api` package. Also, they don't cover more that a basic sanity check that the JS API...
The `make lint` target that runs the linter locally takes around 1.5 minutes. This is inconvenient and affects the productivity of developers.
Presently e2e tests only check error injections due to the difficulties of validating the delay introduced in requests reliably. As a consequence, the definition of delays in the faults is...