This is a known issue with no workaround except switching the cluster to
legacy mode: https://issues.apache.org/jira/browse/FLINK-10135
On 05.09.2018 19:58, Bryant Baltes wrote:
Hello All,
I have recently made the upgrade from Flink 1.3.2 to 1.5.2.
Everything has gone well, except I'm no longer seeing the
numRunningJobs jobmanager metric. We use this for some notifications
and recovery. According to the docs, it should still be available.
Has something changed or maybe I'm missing something?
Thanks!
