someone has tried the miner scavenger in windows?

  • Never mind the error im getting is the pool rejecting it because its too long of a DL/ I have a better DL already.

  • admin

    @energy I'm using NTFS mount points, so 01, 02, 03 .... are all actually directories, not files

  • I just tried it and it scanned 90TB of plotfiles in 140 seconds. Blago JJM scan these same files in 70 seconds.

    Edit: I adjusted the number of threads to be the number of disks and the time was slightly better than 70 sec. So that is a good thing, but nothing significant IMO.

  • @haitch the problem has been mine, I was thinking about my configuration without realizing the others they can have it in a different way.
    I thought it was weird that you put it like that but today I do not have a good day I'm a bit silly

  • @rds Check that you have this properly configured

  • this is great for me, it scans the same time but uses 1/3 of resources of what blago does, now i can game without having these frame drops during mining time lol

  • admin

    @hidevin I want my plotters to finish so I can properly evaluate it ....

    @Energy LOL - no problem. My setup isn't like most others 😉

  • ON MY 4 CORE CPU best performance / efficiency balance:
    worker_thread_count: 2
    reader_thread_count: 2
    Not the fastest option for me i'm afraid.
    This is something for more than 4 core CPU's I'd say 8 cores to keep balance in efficiency but because I do not have such; its only theory.
    I bet servers with 32 cores would benefit.
    That is, 16 8TB drives or plots at once
    I did the following:
    reader_thread_count: 4
    It worked but like Blago it maxed out my cpu - same speed result
    such configuration I would advise those whom want best performance
    I increased reader thread to 6 and it gave a small boost so perhaps it is cache based?

  • I just ran scavenger on my big machine. 288TB 35 seconds with Blago, 25 sec with scavenger.

  • to be a beta the performance does not seem bad