Mark Paluch

Hi Charlie,

you're right. Possible is everything and for sure better than testing after deployment. I'm a fan of fast and simple, therefore, this approach. Using Jersey as in-memory HTTP server is as simple as RESTEasy.

Running jetty/tomcat/wildfly using Maven requires you to change the context in a certain manner. This context change might seem not much, yet it is there. My other point is, your jetty/tomcat/wildfly needs to start up, needs to deploy, and has to shut down (which takes a couple of seconds).

Maven with jetty/tomcat/wildfly takes at least 10x to 20x more time than isolated testing with the in-memory example, and the deployment contains more classes than you need for your test. The use of Mockito (or other mock frameworks) requires Arquillian or something similar. On the other hand, if you start your jetty/tomcat/wildfly once it might be overall faster than starting/stopping a HTTP in-memory server for each unit test case.

Best regards, Mark

Hi Charlie,

you're right. Possible is everything and for sure better than testing after deployment. I'm a fan of fast and simple, therefore, this approach. Using Jersey as in-memory HTTP server is as simple as RESTEasy.

Running jetty/tomcat/wildfly using Maven requires you to change the context in a certain manner. This context change might seem not much, yet it is there. My other point is, your jetty/tomcat/wildfly needs to start up, needs to deploy, and has to shut down (which takes a couple of seconds).

Maven with jetty/tomcat/wildfly takes at least 10x to 20x more time than isolated testing with the in-memory example, and the deployment contains more classes than you need for your test. The use of Mockito (or other mock frameworks) requires Arquillian or something similar. On the other hand, if you start your jetty/tomcat/wildfly once it might be overall faster than starting/stopping a HTTP in-memory server for each unit test case.

Best regards, Mark

Hi Charlie,

you're right. Possible is everything and for sure better than testing after deployment. I'm a fan of fast and simple, therefore, this approach. Using Jersey as in-memory HTTP server is as simple as RESTEasy.

Running jetty/tomcat/wildfly using Maven requires you to change the context in a certain manner. This context change might seem not much, yet it is there. My other point is, your jetty/tomcat/wildfly needs to start up, needs to deploy, and has to shut down (which takes a couple of seconds).

Maven with jetty/tomcat/wildfly takes at least 10x to 20x more time than isolated testing with the in-memory example, and the deployment contains more classes than you need for your test. The use of Mockito (or other mock frameworks) requires Arquillian or something similar. On the other hand, if you start your jetty/tomcat/wildfly once it might be overall faster than starting/stopping a HTTP in-memory server for each unit test case.

Best regards, Mark

Hi Charlie,

you're right. Possible is everything and for sure better than testing after deployment. I'm a fan of fast and simple, therefore, this approach. Using Jersey as in-memory HTTP server is as simple as RESTEasy.

Running jetty/tomcat/wildfly using Maven requires you to change the context in a certain manner. This context change might seem not much, yet it is there. My other point is, your jetty/tomcat/wildfly needs to start up, needs to deploy, and has to shut down (which takes a couple of seconds).

Maven with jetty/tomcat/wildfly takes at least 10x to 20x more time than isolated testing with the in-memory example, and the deployment contains more classes than you need for your test. The use of Mockito (or other mock frameworks) requires Arquillian or something similar. On the other hand, if you start your jetty/tomcat/wildfly once it might be overall faster than starting/stopping a HTTP in-memory server for each unit test case.

Best regards, Mark

Hi Charlie,

you're right. Possible is everything and for sure better than testing after deployment. I'm a fan of fast and simple, therefore, this approach. Using Jersey as in-memory HTTP server is as simple as RESTEasy.

Running jetty/tomcat/wildfly using Maven requires you to change the context in a certain manner. This context change might seem not much, yet it is there. My other point is, your jetty/tomcat/wildfly needs to start up, needs to deploy, and has to shut down (which takes a couple of seconds).

Maven with jetty/tomcat/wildfly takes at least 10x to 20x more time than isolated testing with the in-memory example, and the deployment contains more classes than you need for your test. The use of Mockito (or other mock frameworks) requires Arquillian or something similar. On the other hand, if you start your jetty/tomcat/wildfly once it might be overall faster than starting/stopping a HTTP in-memory server for each unit test case.

Best regards, Mark

Hi Charlie,

you're right. Possible is everything and for sure better than testing after deployment. I'm a fan of fast and simple, therefore, this approach. Using Jersey as in-memory HTTP server is as simple as RESTEasy.

Running jetty/tomcat/wildfly using Maven requires you to change the context in a certain manner. This context change might seem not much, yet it is there. My other point is, your jetty/tomcat/wildfly needs to start up, needs to deploy, and has to shut down (which takes a couple of seconds).

Maven with jetty/tomcat/wildfly takes at least 10x to 20x more time than isolated testing with the in-memory example, and the deployment contains more classes than you need for your test. The use of Mockito (or other mock frameworks) requires Arquillian or something similar. On the other hand, if you start your jetty/tomcat/wildfly once it might be overall faster than starting/stopping a HTTP in-memory server for each unit test case.

Best regards, Mark

Hi Charlie,

you're right. Possible is everything and for sure better than testing after deployment. I'm a fan of fast and simple, therefore, this approach. Using Jersey as in-memory HTTP server is as simple as RESTEasy.

Running jetty/tomcat/wildfly using Maven requires you to change the context in a certain manner. This context change might seem not much, yet it is there. My other point is, your jetty/tomcat/wildfly needs to start up, needs to deploy, and has to shut down (which takes a couple of seconds).

Maven with jetty/tomcat/wildfly takes at least 10x to 20x more time than isolated testing with the in-memory example, and the deployment contains more classes than you need for your test. The use of Mockito (or other mock frameworks) requires Arquillian or something similar. On the other hand, if you start your jetty/tomcat/wildfly once it might be overall faster than starting/stopping a HTTP in-memory server for each unit test case.

Best regards, Mark

Hi Charlie,

you're right. Possible is everything and for sure better than testing after deployment. I'm a fan of fast and simple, therefore, this approach. Using Jersey as in-memory HTTP server is as simple as RESTEasy.

Running jetty/tomcat/wildfly using Maven requires you to change the context in a certain manner. This context change might seem not much, yet it is there. My other point is, your jetty/tomcat/wildfly needs to start up, needs to deploy, and has to shut down (which takes a couple of seconds).

Maven with jetty/tomcat/wildfly takes at least 10x to 20x more time than isolated testing with the in-memory example, and the deployment contains more classes than you need for your test. The use of Mockito (or other mock frameworks) requires Arquillian or something similar. On the other hand, if you start your jetty/tomcat/wildfly once it might be overall faster than starting/stopping a HTTP in-memory server for each unit test case.

Best regards, Mark