RapidMiner generates the Blue Screen of Death

SGolbertSGolbert RapidMiner Certified Analyst, Member Posts: 342   Unicorn
edited August 2019 in Help
Hi all,

I wanted to bring the attention to a long lasting problem with RM that I haven't seen reports or developments on. The problem concerns RM Studio running on Windows (it should happen in Linux too though), and involves RM crashing the system when using too much memory.

It has happen to me also when using good machines (32 GB RAM), specially when running something in AutoModel in the background. I thougth the JVM was inmune to these issues, but that is clearly not the case.

Is the software development team aware of the problem or doing something to prevent it? The ideal solution would be to abort the process (one of them, in case multiple process are running). At the worst close all process and RM itself. But NEVER crash my OS!


Best Answers


  • SGolbertSGolbert RapidMiner Certified Analyst, Member Posts: 342   Unicorn

    That's a good work around, if it works. I still wonder why does the OS crash. Isn't it possible to use virtual memory?

  • varunm1varunm1 Moderator, Member Posts: 1,069   Unicorn
    Actually, it should throw an outofmemory error and stop. I am not sure why its crashing OS. May be @Marco_Boeck have some insights.
  • Marco_BoeckMarco_Boeck Team Lead Software Engineering Administrator, Moderator, Employee, Member, University Professor Posts: 1,887   RM Engineering

    To be perfectly honest here, I'd love it if I could create bluescreens by coding regular Java :# , but the truth is, I cannot. If something happens, it would be outside of our control, like a result of some unfortunate combination of the Java VirtualMachine and other external factors like @hughesfleming68 mentioned above.

    Therefore there is pretty much nothing we can do on our end.

Sign In or Register to comment.