Relative Content

Tag Archive for apache-flinkflink-streamingflink-sqlflink-cepflink-batch

Ephemeral storage bloat when submitting multiple jobs to flink session cluster

I’m try to run a flink session cluster and submit a lot of small jobs on it.
My jar is configurable on what processing should happen, where the data is and to where the processed data should be loaded.
The problem i have is that when i try to submit more than 5 jobs at once the cluster struggles until it fails.
The blob server takes giga bytes of storage for jobs that are very small and simple.
And the job manager is very overloaded with planning all jobs which are pretty much the same job with different configuration.
Is there a way to handle my problem better with flink? Is flink not the solution?

Ephemeral storage bloat when submitting multiple jobs to flink session cluster

I’m try to run a flink session cluster and submit a lot of small jobs on it.
My jar is configurable on what processing should happen, where the data is and to where the processed data should be loaded.
The problem i have is that when i try to submit more than 5 jobs at once the cluster struggles until it fails.
The blob server takes giga bytes of storage for jobs that are very small and simple.
And the job manager is very overloaded with planning all jobs which are pretty much the same job with different configuration.
Is there a way to handle my problem better with flink? Is flink not the solution?