Exception in thread "Thread-30" java.lang.OutOfMemoryError: Direct buffer memory

Khalid538

I am getting this error when one of our Data Scientist is running a Job. Currently i have assigned backed.xmx memory to 8G planning to increase it to 20 GB for larger data processing.

Is this the solution ?.

---

Operating system used: Cent OS

CoreyS

Hi @Khalid538
and welcome to the Dataiku Community. While you wait for a more detailed response, please note that for large production instances, we recommend allocating 12 to 20 GB of memory for the backend. More information can be found in our documentation:

• Tuning and controlling memory usage
• Understanding and tracking DSS processes
• Memory Optimization Tips: Backend, Python/R, Spark jobs (Knowledge Base)

I hope this helps!