Hi Alexandru As for 2, generally speaking the number of required slots depends on number of slot sharing groups. By default all operators belong to the default slot sharing group, that means a job requires as many slots as maximal parallelism in the job. More on the distributed runtime you can read here[1]
As for 1 I cc'ed Gary and Till who might better answer your question. [1] https://ci.apache.org/projects/flink/flink-docs-release-1.7/concepts/runtime.html#task-slots-and-resources Best, Dawid On 14/01/2019 15:26, Alexandru Gutan wrote: > Hi everyone! > > 1. Is there a way to increase the parallelism (e.g. through REST) of > some operators in a job without re-deploying the job? I found this > <https://stackoverflow.com/questions/50719147/apache-flink-guideliness-for-setting-parallelism> > answer which mentions scaling at runtime on Yarn/Mesos. Is it > possible? How? Support for Kubernetes? > 2. What happens when the number of parallel operator instances exceeds > the number of task slots? For example: a job with a source > (parallelism 3), a map (parallelism 8), a sink (parallelism 3), total > of *14* operator instances and a setup with *8* task slots. Will the > operators get chained? What if I disable operator chaining? > > Thank you!
signature.asc
Description: OpenPGP digital signature
