Hi Curtis, we implemented this today. But without a REST-Interface. We transfer out artifacts and a script with a scp call from out Bamboo server and execute the script. This script kills the yarn application, start a new flink application in yarn and submit all routes to the cluster. Works pretty well... ;) Just in case... Greets Dominique Von meinem Samsung Gerät gesendet. -------- Ursprüngliche Nachricht --------
Von: Curtis Wilde <[hidden email]> Datum: 22.09.16 18:49 (GMT+01:00) An: [hidden email] Betreff: REST Interface to JobManager I would like to be able to use Jenkins to deploy jobs to Flink. I’ve seen talk of a REST interface that might allow me to do this
https://issues.apache.org/jira/browse/FLINK-1228 Is there any documentation around this feature? |
It looks like I was able to finally find what I needed. WebRuntimeMonitor.java contains the routes, with a little fiddling in the browser I think I can make the rest happen. https://github.com/apache/flink/blob/master/flink-runtime-web/src/main/java/org/apache/flink/runtime/webmonitor/WebRuntimeMonitor.java From:
Dominique Rondé <[hidden email]> Hi Curtis, we implemented this today. But without a REST-Interface. We transfer out artifacts and a script with a scp call from out Bamboo server and execute the script. This script kills the yarn application, start a new flink application in yarn
and submit all routes to the cluster. Works pretty well... ;) Just in case... Greets Dominique Von meinem Samsung Gerät gesendet. -------- Ursprüngliche Nachricht -------- I would like to be able to use Jenkins to deploy jobs to Flink. I’ve seen talk of a REST interface that might allow me to do this
https://issues.apache.org/jira/browse/FLINK-1228 Is there any documentation around this feature? |
Hi, We went a different way and used the same underlying ClusterClient classes from the command line tool, to connect to the JobManager and submit JARs that way. From the YARN RM address, we search for the Flink application and use the YarnClusterClient to get the connection details from the discovered application id. We then use the YarnClusterClient to submit the job and later get progress updates. Thanks, Paul On 22 September 2016 at 20:24, Curtis Wilde <[hidden email]> wrote:
|
Free forum by Nabble | Edit this page |