The most probable cause of this is that your cluster is overburdened either with other users, or with other processes. Check with your Xcalar Admin to see if other users may be hogging system resources, or check out the system itself to ensure that there is no other process hogging up the memory.
If that doesn't fix your solution, here's some other common issues:
- Make sure your streaming UDF is efficient; search for Python best practices for performance
- Make sure your storage is well-tuned
- Ensure your system has been set up according to the documented minimum system requirements