Maintenance tips for your hard drives and plots Earthlings!



  • @ZapbuzZ,

    I played around with the cache size on the Blago CPU miner awhile back. I saw no significant difference in read times unless the cache was squashed towards zero. Super large caches actually resulted in slightly longer scan times. I ended up resetting back to the default value, 100,000 IIRC.



  • @rds i used the non avx version. I also used the larger buffer when my plot started to really grind down to 1/3 its usual speed. I eventually used the plot optimiser. But the larger buffer kept the cpu fed its maximum for many weeks faster. I assume you had new plots? i found that having SATA3 drives plugged into SATA 2 or 1 motherboards/controllers can be a bottleneck.



  • @ZapbuzZ,

    I was using avx, and all my drives are on USB 3.0 through ganged hubs.



  • @rds i'd be very helpful if you could share with me, some photos of your equipment setup that I may be able to offer suggestions. It may cost some money, however, but first thing that comes to my mind is lots of hubs less throughput. But before i explain further i need to see what you use. I don't mind private messages of your photos if its possible. (your pc including the motherboard, connections to hubs)



  • Oh windows users: you can configure internal hard disks to join together to make 1 massive disk out of smaller disks for example 4 x 4gb = 1 x 16 gb disk. the method is called volume spanning and is available with disk management through administrative tools;, and doesn't matter what size or speed your disks are/ this is useful to conserve drive letters and to plot some very huge plots meaning less plot names to process. windows 10's control panel has a function called storage spaces that can even add external drives. i have a 7200rpm 3TB internal and a 4GB 4200rpm disk spanned together no problems and I can later add more whenever I want to.



  • @ZapbuzZ said in Maintenance tips for your hard drives and plots Earthlings!:

    Oh windows users: you can configure internal hard disks to join together to make 1 massive disk out of smaller disks for example 4 x 4gb = 1 x 16 gb disk. the method is called volume spanning and is available with disk management through administrative tools;, and doesn't matter what size or speed your disks are/ this is useful to conserve drive letters and to plot some very huge plots meaning less plot names to process. windows 10's control panel has a function called storage spaces that can even add external drives. i have a 7200rpm 3TB internal and a 4GB 4200rpm disk spanned together no problems and I can later add more whenever I want to.

    There is just one problem with this I believe, and that is if one of your disks dies that is spanned together with the others, I think the whole spanned volume is corrupted, but I am not sure. You can use disk manager to delete volume, then create a new simple volume in an empty folder on C for example. Then each volume is tied to one disk, and you can have as many as you like.



  • @ZapbuzZ said in Maintenance tips for your hard drives and plots Earthlings!:

    @rds i'd be very helpful if you could share with me, some photos of your equipment setup that I may be able to offer suggestions. It may cost some money, however, but first thing that comes to my mind is lots of hubs less throughput. But before i explain further i need to see what you use. I don't mind private messages of your photos if its possible. (your pc including the motherboard, connections to hubs)

    My throughput is 220 MB/s. My scans are done in about 60 seconds.



  • :)



  • thats a good speed though and would match a high en gpu/cpu too @rds



  • the miners in the wallet application can be manually tweaked. The best way to do that is to copy the miner of choice (after it has been setup through the app) and launched from the miner itself as the app (wallet AIO) overwrites the miners config within the app. Biago's miner has switches that the AIO wallet doesn't tweak.


Log in to reply
 

Looks like your connection to Burst - Efficient HDD Mining was lost, please wait while we try to reconnect.