You can also set up a runtime image that fetches and runs the fat jar from a known URL (versioned) and then your base image wouldn't change, just the env var pointing to the fat jar. We don't have a standard name for that, but it's basically a lazy binary build.
On Jun 9, 2017, at 6:51 AM, Jens Geiregat <[email protected]> wrote: Hi Marcin, I asked part 2 of your question here some time ago. Check out Ben Parees' answer and my answer to him here: http://lists.openshift.redhat.com/openshift-archives/users/2017-February/msg00065.html We've been using the binary build since then and it works great. Kind Regards, Jens On Fri, Jun 9, 2017 at 11:40 AM, Marcin Zajączkowski <[email protected]> wrote: > Hi. I'm evaluating OpenShift 3 to use it in my CI/CD infrastructure (I > have just some past experience with OpenShift 2). Currently Jenkins builds > fat-jars (Spring Boot) and deploy it to Nexus. Later on in the pipeline > they are deployed to test environment and some smoke testing is run. I > would like to have those µservices deployed in OpenShift to have > better resource utilization - there are ~200 different µservices and > usually there are 0 to 60 deployed on test environment in the > same time. Having one big instance or manual sharding is quite > problematic. fat-jars are already build and I just need to deploy it (no > pipelines on the OS part, etc.). Therefore, I have two questions. > > Technically, currently I have a Git repo with Dockerfile extending a > fabric8/java-alpine-openjdk8-jdk image, exposing a port and downloading a > JAR from Nexus (built with a docker strategy). It works in general, > however, it seems to be quite generic solution. I could probably also > create a custom imagestream based on aforementioned image and create > different services with a JAR passed as an environment. > > > 1. What do you thing in general about the idea to use OpenShift just for > scaling up and down a number of total service instances (to reduce resource > utilization)? Can OpenShift be easily used for that or maybe you see some > caveats? > 2. What is the easiest (or recommended) way just deploy a binary fat-jar > to OpenShift (as described in my case)? > > Marcin > > -- > http://blog.solidsoft.info/ - Working code is not enough > > > > _______________________________________________ > users mailing list > [email protected] > http://lists.openshift.redhat.com/openshiftmm/listinfo/users > _______________________________________________ users mailing list [email protected] http://lists.openshift.redhat.com/openshiftmm/listinfo/users
_______________________________________________ users mailing list [email protected] http://lists.openshift.redhat.com/openshiftmm/listinfo/users
