Is anyone actively working on direct Kubernetes support?
I'd be excited to see this get in sooner rather than later, I'd
be happy to start a PR.
On 3/22/18 10:37 AM, Till Rohrmann
wrote:
Hi Edward and Eron,
you're ri
Hi Edward and Eron,
you're right that there is currently no JobClusterEntrypoint implementation
for Kubernetes. How this entrypoint looks like mostly depends on how the
job is stored and retrieved. There are multiple ways conceivable:
- The entrypoint connects to an external system from which it
It would be helpful to expand on how, in job mode, the job graph would be
produced. The phrase 'which contains the single job you want to execute'
has a few meanings; I believe Till means a serialized job graph, not an
executable JAR w/ main method. Till is that correct?
On Tue, Mar 20, 2018 at
Hi Till,
Thanks for the information. We are using the session cluster and is working
quite good, but we would like to benefit from the new approach of per-job
mode in order to have a better control over the jobs that are running on the
cluster.
I took a look to the YarnJobClusterEntrypoint and I
Hi Edward,
you're right that Flink's Kubernetes documentation has not been updated
with respect to Flip-6. This will be one of the tasks during the Flink 1.5
release testing and is still pending.
A Flink cluster can be run in two modes: session mode vs per-job mode. The
former starts a cluster to
Hello,
Currently I have a Flink 1.4 cluster running on kubernetes based on the
configuration describe on
https://ci.apache.org/projects/flink/flink-docs-release-1.4/ops/deployment/kubernetes.html
with additional config for HA with Zookeeper.
With this I have several Taskmanagers, a single Jobmana
