Forum Discussion

  • Mikael's avatar
    Mikael
    Brass Contributor

    Me again!


    I've searched high and low, trying to educate myself on the subject. Not much found on the behavior I'm experience.

    Bugging Co-pilot gives very erratic answers. It links and explains the Deduplication function in Windows not ReFS dedup...

    I rephrased and ask especially for Refs and got this answer.

    For ReFS (Resilient File System), there are no specific memory requirements tied directly to the amount of data, as ReFS is designed to handle large volumes efficiently without significant memory overhead. ReFS focuses on providing data integrity, availability, and scalability without the need for extensive memory resources

    So, again. No explanation why ReFS Dedup just eats up all memory and crashes the server...

     

    Brought some information here from the other post.

    This is what happen when a start, and in the particular situation here I stopped it before the server exhausted it's memory.

    Start-ReFSDedupJob -Volume L: -CpuPercentage 30
    Stop-ReFSDedupJob -Volume L:

    The argument -Fullrun results in the same behavior. 

    L:, It's 12TB volume allocated with 10,5TB of data.

     

Resources