• Posts

  • Joined

  • Last visited

About mdrodge

  • Birthday February 18


  • Member Title


  • Gender
  • URL

Recent Profile Visitors

987 profile views

mdrodge's Achievements


Explorer (4/14)



  1. I went through this about 3 years or so ago. (i have an expensive day rate and a cheaper night rate for power) I sold my 2990wx and bought a 15w intel itx board and set the hdd's to power down after 15mins. ("main" server is barley measurable at idle) I build a second system with a 5700x and underclocked to 3ghz, I have that system running a script that turns off resilio_sync so that Handbrake gets no new jobs during the day when power is expensive and it gets to work at night. I also got a few solar panels (inverter, transfer switch etc) and put them on my shed and ran a cable to my systems. Also also i got a couple of GPU's for the render box and set up HiveOS as a VM with the GPU's on passthrough so i could mine ETH (plus a lot more mining stuff around the house that i won't go into) but basically my servers MAKE money Instead of Costing. How much they make depends on how much sun there is, how many handbrake jobs they get and the price of ETH but it seems to be between £1 - £2 a day profit. I some of these things were not an option to me i would just run all my compute on %100 solar and aim for ultra low power from the grid (basically I'd have my render box on a separate circuit that was only powered by the sun and I'd have to try and make do with 3 hours (average) compute a day)
  2. A couple of us where working out the same issue on discord today. (I'll tell you what they told me) in the cactus image console type : vi /machinaris/scripts/forks/ this should bring up a file, there seems to be an extra If command... delete it ( a line with if and nothing else ) then move to the bottom of the page and type :wq (to save)
  3. If that happens it's normally because unraid had no cpu time available, when setting up any vm (or docker image or anything else) always leave at least core 0 free for unraid to use. (if your moving lots of files about you might need a few cores. ...Once the CPU goes down it might come back but remember that it's not fixed and sooner or later expect a hard lock up. (i've even corrupted unraid USB's like this so i have learned the hard way) (though I've also had an issue where the system hard locks because I'd hooked up to many gpu's for that mobo's bios (gigabyte b550) (that board won't run 4 gpu's even on bare metal and 2 is the max in a vm on that system)
  4. It seems to mess up the network.cnfg file some how. It can't get diagnostic on it because it's offline until the network.cnfg has been removed and the lan cables swapped and new ip issued. I was told that this is a greater Linux issue though, something about the driver from Realtek apparently.
  5. rc5 is supposed to fix the issue. I'll test it later (when the kids are asleep) if i don't comment in 24 hours then assume it's working or at least in testing but not broken enough to come back and update. Update..... still the same. no fix.
  6. Thanks for catching me up. I did see that people are reporting this in Linux forums to so hopefully....... I was wondered if this was the case. Though if nothing else I'd just like the info up here somewhere because once this hits stable there is going to be a few annoyed people wondering what's going on (unless it's been mentioned at least)
  7. My one works in Windows and on the Stable but won't show up with RC3 or higher.
  8. So Realtek RTL8156 and RTL8125 based NIC's both broken by rc4 @limetechUrgent 😃
  9. There are a few of us that have reported 2.5gb network issues with RC4 but as yet no one has acknowledged this. So far it seems tied to the RTL8156, is this your chipset?
  10. Same problem, thread started here (It is just me so far) I really need a long term solution (not just don't update)
  11. After you spent 4.5gb of vram on ETHash that would only leave you with 1.5gb for transcodeing anyway. (And there is the whole needing to unlock the driver block thing) I used to do plex with a p2000 but I sold it off when I realised I could get the decent transcodeing with an intel igpu. Cpu transcodeing isn't the end of the world anyway. (But those intel igpus are very efficient)
  12. the guide t-rex main support forum is about as good as it gets. did you have a specific question (personally i think the docker of t-rex has a horrible ui so i won't use it but i have done) (also i didn't get very good hashrate with it) I've rewritten this 4 times now but i'm not really sure what you want to know, I'd probably try and push you in a different direction but i'm not sure what you have and what your trying to do. I'm happy to guide you though if you need, i am sure we can figure it out.