That's true if you use RDP because RDP unloads video drivers when you connect. Try using VNC, or better yet, Teamviewer. I've had great success with Teamviewer. You can set it to automatically connect/login and you don't have to be on the same LAN to connect to the box. Plus it has a free Android app.
+1 and there's a free team viewer app for iPhone. 
Topic: 2 Nodes @ 100% Load for 2 Months | No Humans | Max Uptime! (Read 1661 times)
BAMT 0.4B will be perfect!
If you want to try linux, here are some hints...
Using WiFi (Cheap Router & Adapters)
Not sure what the pricing or coverage is like in your region, but using a mobile broadband(3G or what ever might be available) as a backup/fail over connection might be of interest if your internet connection have any history of not being reliable.
No UPS
Set the bios to "Start" if power is lost and then restored.
Summer
Code a watcher script that downclocks the card as temperatures go above predefined values, and overclocks in steps or full when temperature is within safe levels.
You might want to go extreme and use the motherboard temperature as an indicator of the room temperature so that it doesn't hang the GPU almost directly after startup if too aggressive overclocking is applied at start.
No Config to continue correctly if system reboots
Create a simple startup-script
No One to check on it
Code a simple monitor script that checks for GPUs that are hung etc, and coldreboots the rig if hungs are detected.
If your motherboards support IPMI or similar, then the rigs can monitor each other, and take action accordingly, such as reboot etc.
A managed PDU could be interesting as well, have the rigs monitor each other, if problems occur or a dead response and the system doesn't want to reboot, the other rig tells the PDU to power-cycle the troubling rig.
No Communication
same as above
No Access to a computer
same as above
Using WiFi (Cheap Router & Adapters)
Not sure what the pricing or coverage is like in your region, but using a mobile broadband(3G or what ever might be available) as a backup/fail over connection might be of interest if your internet connection have any history of not being reliable.
No UPS
Set the bios to "Start" if power is lost and then restored.
Summer
Code a watcher script that downclocks the card as temperatures go above predefined values, and overclocks in steps or full when temperature is within safe levels.
You might want to go extreme and use the motherboard temperature as an indicator of the room temperature so that it doesn't hang the GPU almost directly after startup if too aggressive overclocking is applied at start.
No Config to continue correctly if system reboots
Create a simple startup-script
No One to check on it
Code a simple monitor script that checks for GPUs that are hung etc, and coldreboots the rig if hungs are detected.
If your motherboards support IPMI or similar, then the rigs can monitor each other, and take action accordingly, such as reboot etc.
A managed PDU could be interesting as well, have the rigs monitor each other, if problems occur or a dead response and the system doesn't want to reboot, the other rig tells the PDU to power-cycle the troubling rig.
No Communication
same as above
No Access to a computer
same as above
heres a guide i used https://bitcointalksearch.org/topic/guidehow-to-setup-an-automated-headless-linuxcoin-mining-rig-watchdog-29021
it mostly worked for me.
previously i had windows but i felt it was bloated and kept crashing. linux is more of a server os built for 24/7 stuff.
it mostly worked for me.
previously i had windows but i felt it was bloated and kept crashing. linux is more of a server os built for 24/7 stuff.
You planning on using Linux or Windows? If it's linux, then flashing the card bios is probably your best bet. Windows, I'd use something like MSI Afterburner or Trixx and set it to autostart. Though since these boxes have to run without any supervision, I'd go easy on any overclocks just to keep the cards as stable as possible.
You may want to set the box to automatically reboot itself every couple of days just to clear up and hung up miners, etc. If you can set this in the BIOS, all the better in case the OS is hung up.
As for the power question, you should be good with either PSU on either machine. Both those CPU's run at about 95W. As for the cards, the 5850 is about 150W, 5870 about 200W and the 6950's are about 175W each. (If you overclock a lot, that'll go up). So, the 850 will have more than enough headroom for either setup. You could even underclock the CPU's a bit if you want to save some power.
Couple other things I can think of. Make sure you have your miners setup with a backup pool/miner in case the primary one goes down.
Also, you might want to get this up and operational a week or so before you plan on heading out and let it run on it's own so you can shake out any major problems.
You may want to set the box to automatically reboot itself every couple of days just to clear up and hung up miners, etc. If you can set this in the BIOS, all the better in case the OS is hung up.
As for the power question, you should be good with either PSU on either machine. Both those CPU's run at about 95W. As for the cards, the 5850 is about 150W, 5870 about 200W and the 6950's are about 175W each. (If you overclock a lot, that'll go up). So, the 850 will have more than enough headroom for either setup. You could even underclock the CPU's a bit if you want to save some power.
Couple other things I can think of. Make sure you have your miners setup with a backup pool/miner in case the primary one goes down.
Also, you might want to get this up and operational a week or so before you plan on heading out and let it run on it's own so you can shake out any major problems.
Thanks I am doing that now, I have until the 15th of August, I am about 90% done building the second node, kind of disappointed with the two 6950's performance to price ratio since I paid $260 for each and get the same performance as my $150 5850 Extreme and way less performance than my 5870 Vapor X (best single gpu mining card in my opinion). I picked it up for $220 before mining really for popular and I get 440 mhash/s out of it stable in the middle of summer. When I put the case fans in hopefully I will get a nice temp drop and be able to push the clocks higher, but these dual fan cooling systems seem inferior to the ones used on the sapphire extreme and vapor x cards.
I'm pretty sure if you remotely connect to the computer, you will not be able to start the miner (or at least it wouldn't work for me), the cards won't show up.
What I did (make sure your cards are stable though) is use Task Scheduler, and run a shutdown restart command daily, followed by afterburner, and poclbm. Works great for me.
What I did (make sure your cards are stable though) is use Task Scheduler, and run a shutdown restart command daily, followed by afterburner, and poclbm. Works great for me.
That's true if you use RDP because RDP unloads video drivers when you connect. Try using VNC, or better yet, Teamviewer. I've had great success with Teamviewer. You can set it to automatically connect/login and you don't have to be on the same LAN to connect to the box. Plus it has a free Android app.
You planning on using Linux or Windows? If it's linux, then flashing the card bios is probably your best bet. Windows, I'd use something like MSI Afterburner or Trixx and set it to autostart. Though since these boxes have to run without any supervision, I'd go easy on any overclocks just to keep the cards as stable as possible.
You may want to set the box to automatically reboot itself every couple of days just to clear up and hung up miners, etc. If you can set this in the BIOS, all the better in case the OS is hung up.
As for the power question, you should be good with either PSU on either machine. Both those CPU's run at about 95W. As for the cards, the 5850 is about 150W, 5870 about 200W and the 6950's are about 175W each. (If you overclock a lot, that'll go up). So, the 850 will have more than enough headroom for either setup. You could even underclock the CPU's a bit if you want to save some power.
Couple other things I can think of. Make sure you have your miners setup with a backup pool/miner in case the primary one goes down.
Also, you might want to get this up and operational a week or so before you plan on heading out and let it run on it's own so you can shake out any major problems.
You may want to set the box to automatically reboot itself every couple of days just to clear up and hung up miners, etc. If you can set this in the BIOS, all the better in case the OS is hung up.
As for the power question, you should be good with either PSU on either machine. Both those CPU's run at about 95W. As for the cards, the 5850 is about 150W, 5870 about 200W and the 6950's are about 175W each. (If you overclock a lot, that'll go up). So, the 850 will have more than enough headroom for either setup. You could even underclock the CPU's a bit if you want to save some power.
Couple other things I can think of. Make sure you have your miners setup with a backup pool/miner in case the primary one goes down.
Also, you might want to get this up and operational a week or so before you plan on heading out and let it run on it's own so you can shake out any major problems.
I'm pretty sure if you remotely connect to the computer, you will not be able to start the miner (or at least it wouldn't work for me), the cards won't show up.
What I did (make sure your cards are stable though) is use Task Scheduler, and run a shutdown restart command daily, followed by afterburner, and poclbm. Works great for me.
What I did (make sure your cards are stable though) is use Task Scheduler, and run a shutdown restart command daily, followed by afterburner, and poclbm. Works great for me.
You can easily set up your miners to auto start on login through either the miner itself, a windows task or even showing the miner shortcut or a batch file into the Startup folder. Of course that's using windows, and considering you want to keep the machine running 2 months straight, you might want to look into linux. Either way, to keep the machine stable, you may want to go easy on the overclocking.
Also, most machines nowadays have a BIOS setting in the power settings that you can tell the PC what to do when it recovers from a power outtage. There should be an option to turn the machine on when power returns. There are also devices out there you can buy to cycle power remotely if you really want to deal with that in case windows locks up and isn't responsive.
Finally, assuming you have internet connectivity for those machines, just use Teamviewer or VNC to remotely connect to them.
Also, most machines nowadays have a BIOS setting in the power settings that you can tell the PC what to do when it recovers from a power outtage. There should be an option to turn the machine on when power returns. There are also devices out there you can buy to cycle power remotely if you really want to deal with that in case windows locks up and isn't responsive.
Finally, assuming you have internet connectivity for those machines, just use Teamviewer or VNC to remotely connect to them.
All that sounds great and thank you, I will not have access to a pc or cell phone at all so VNC is out, they will be on their own. What about having the clocks set back to what I need when they restart. Would it be smarter to flash the bios on the cards to keep them at the right clocks or just have the software change it at startup? Yes I have that bios option set already.
Any other advice will be much appreciated , I want to try to cover myself as much as possible.
Also kind of off topic hardware question...
Which of these 2 configs use more power is everything else was identical but these components
Intel Core2Quad Q9550 @ 2.8 + 5870 Vapor X + 5850 eXtreme
Intel Core i5 2500 @ 3.3 + 6950 TwinFrozr + 6950 Dirt3 Edition
Everything else will be pretty much the same. The reason I ask is I have 2 Corsair PSU's, one is Corsair TX850 and other is HX1000. Does it matter which gets which?
You can easily set up your miners to auto start on login through either the miner itself, a windows task or even showing the miner shortcut or a batch file into the Startup folder. Of course that's using windows, and considering you want to keep the machine running 2 months straight, you might want to look into linux. Either way, to keep the machine stable, you may want to go easy on the overclocking.
Also, most machines nowadays have a BIOS setting in the power settings that you can tell the PC what to do when it recovers from a power outtage. There should be an option to turn the machine on when power returns. There are also devices out there you can buy to cycle power remotely if you really want to deal with that in case windows locks up and isn't responsive.
Finally, assuming you have internet connectivity for those machines, just use Teamviewer or VNC to remotely connect to them.
Also, most machines nowadays have a BIOS setting in the power settings that you can tell the PC what to do when it recovers from a power outtage. There should be an option to turn the machine on when power returns. There are also devices out there you can buy to cycle power remotely if you really want to deal with that in case windows locks up and isn't responsive.
Finally, assuming you have internet connectivity for those machines, just use Teamviewer or VNC to remotely connect to them.
My goal is pretty simple to understand , not so simple to achieve.
I need two nodes to run for two months without any help from anything but themselves.
I have around 2 weeks to figure this out. Hash Node 2 is being built right now. I have had success keeping HN1 running for over 5 days until I stopped them for some reason I can't remember. I use Windows 7 x64 with all the enhanced GUI features off in combination with the PCLBM GUI with excellent results. I will be running two separate nodes, one with 6950x2 and one with a 5870/5850. Both intel based, one is P45 other is Z68.
Hear are my main concerns,
Using WiFi (Cheap Router & Adapters)
No UPS
Summer
No Config to continue correctly if system reboots
No One to check on it
No Communication
No Access to a computer
Plus any other variables...
Shit late for work, be back later, thanks in advance for any advice, I will try to logon at work...
I need two nodes to run for two months without any help from anything but themselves.
I have around 2 weeks to figure this out. Hash Node 2 is being built right now. I have had success keeping HN1 running for over 5 days until I stopped them for some reason I can't remember. I use Windows 7 x64 with all the enhanced GUI features off in combination with the PCLBM GUI with excellent results. I will be running two separate nodes, one with 6950x2 and one with a 5870/5850. Both intel based, one is P45 other is Z68.
Hear are my main concerns,
Using WiFi (Cheap Router & Adapters)
No UPS
Summer
No Config to continue correctly if system reboots
No One to check on it
No Communication
No Access to a computer
Plus any other variables...
Shit late for work, be back later, thanks in advance for any advice, I will try to logon at work...
Jump to: