LinkedIn
Twitter
Hi Folks! I would like to know if anyone as any idea of what could be happening here. I have a streaming job which works fine. However, memory utilziation on worker keeps growing forever, until the point it reaches the max capacity of the worker and then the worker is shut down and a new one is created. Pipeline does not fail, on these moments there is a latency in processing messages which can get to around 10 minutes, but then wverything returns to normal. Any idea what could be bad to generate this behaviour on memory utilization? It looks like memory is never flushed from processed messages.
