100TB Burst rig questions



  • @Evo Did not know that! Thanks very much for educating me :D



  • @haitch, that's pretty much what I assumed. Good of you to offer to advise on choices. Thanks. BTW, regarding wives, I think I have a winner: she's had to have noticed that the line of drives keeps getting longer and longer, but she hasn't asked once about how much I've spent. She will ask "how much did you make today". I tell her. I don't tell her how long it's going to be before I hit my ROI. Fingers crossed that things stay as they are!

    @IncludeBeer, no problem! I've had pretty good luck plotting that way. As I said, I do keep my files under 3TB each. For some reason the damn plotter, at least on my machine, will arbitrarily lock up on the big 6TB to 8TB plots.


  • admin

    @Evo re: lock ups - if you have a decent CPU, use Xplotter. It can resume crashed/halted plots.



  • @haitch said in 100TB Burst rig questions:

    @Evo re: lock ups - if you have a decent CPU, use Xplotter. It can resume crashed/halted plots.

    @haitch, yeah, I'll have to give it a try. Right now I'm almost (one left) done plotting a batch of 8TB SMRs. That will get me North of 100TB, which was my goal. Won't be plotting any more till the next machine, whatever that turns out to be.

    Had a thought after the my previous reply...if the servers can be so tricky for mere mortals like me to get right, what about Frankensteining a decent consumer mobo/processor with one or two GPUs and one or two SATA/SAS expansion cards into one of those 24 bay chassis? How would you imagine a hybrid like this might compare to the all server grade rig? Just curious.


  • admin

    @Evo You could frankenstein one, but you'll still need the right parts - the server chassis mainly use SAS expanders or direct attach connectors, and a lot of the backplanes are limited to 2TB or less drives, so you need to understand the specs of the back plane, and the corresponding capabilities of the controller you're using to drive it. Is it standard SAS SFF8087, or proprietary, X to SFF8087 ?

    Basically it comes down to high drive volume miners require good knowledge or good advice.



  • @haitch said in 100TB Burst rig questions:

    @Evo You could frankenstein one, but you'll still need the right parts - the server chassis mainly use SAS expanders or direct attach connectors, and a lot of the backplanes are limited to 2TB or less drives, so you need to understand the specs of the back plane, and the corresponding capabilities of the controller you're using to drive it. Is it standard SAS SFF8087, or proprietary, X to SFF8087 ?

    Basically it comes down to high drive volume miners require good knowledge or good advice.

    @haitch, it'll have to be good advice! We'll chat down the road.


  • admin

    @Evo No problem, happy to advise anyone on a build - made my own trip ups, and I'm supposed to know what I'm doing ......
    live and learn.



  • @haitch said in 100TB Burst rig questions:

    @manfromafar If they're not new, bitch to eBay .....

    @Dtrade16 Motherboard NIC, and proprietary Power supplies with the chassis. Current rig has dual/redundant 800W PSU's.

    Nice. How do you deal with the noise? Are the fans very loud?


  • admin

    @Dtrade16 The default rig is very loud. I have in-line resistor extensions to slow down the fans, which gets the noise down to a tolerable level. But as it's sitting 3 feet from me, it's not quiet enough yet. Thinking of putting it in our lab at work.



  • @haitch said in 100TB Burst rig questions:

    @Dtrade16 The default rig is very loud. I have in-line resistor extensions to slow down the fans, which gets the noise down to a tolerable level. But as it's sitting 3 feet from me, it's not quiet enough yet. Thinking of putting it in our lab at work.

    Nice, yah that might be a better solution. Trying to get up and running here with this loner computer. So many driver issues. Hate installing drivers lol.


Log in to reply
 

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