{"id":29,"date":"2017-04-12T22:55:48","date_gmt":"2017-04-12T22:55:48","guid":{"rendered":"http:\/\/www.matez.de\/?p=29"},"modified":"2017-04-12T22:55:48","modified_gmt":"2017-04-12T22:55:48","slug":"exploring-feign-retrying","status":"publish","type":"post","link":"https:\/\/www.matez.de\/index.php\/2017\/04\/12\/exploring-feign-retrying\/","title":{"rendered":"Exploring Feign – Retrying"},"content":{"rendered":"

Feign<\/a> is a library, which makes it easier to implement a http client. Recently more and more people start writing http clients, because they are creating microservices which communicate with http protocol. So there are all sorts of libraries supporting this task like Jersey, Resteasy and others – and there is Feign.<\/p>\n

Today I do not want to explain the basic functionality, this is all done on the Readme <\/a>page itself. Today I want to get into the details of a feature, which becomes more and more important, because in modern distributed systems, you want to have resilient behaviour, which means that you want to design your service in the way, that it can handle unexpected situations without noticing on user’s site. For example an API you are calling is not reachable at the moment, the request times out or the requested resource is not yet available. To solve this issue, you need to apply a retry pattern, so that you increase the chance that the service request is successfull after the first, the second or the nth attempt.<\/p>\n

What most developers don’t know, Feign has a default retryer built-in.<\/p>\n

Now I show\u00a0a few code examples, what you can expect from this feature. What I am showing are junit tests with a client mock, so that we are able to stub certain errors and verify, how many retries have been made.<\/p>\n

Case 1) Success<\/h2>\n

no retry needed.<\/p>\n\n

\n
\n
\n \n
\n
\n \n
\n\n \n
\n\n