 Show Posts
|
|
Pages: [1] 2 3 4 »
|
I have not been able to connect to mint.bitminter.com since sometime this morning, directly with cgminer or with a stratum proxy. Trying even to telnet directly to mint.bitminter.com on 3333, 8332, 80, or 5050 results in a connection timeout. I have reloaded the page at https://bitminter.com many times without any problems, although the live stats popup (on the "me" tab) shows no gauge image. Is there a problem? The site seems to be indicating other people are submitting work... *note this is all on a setup that has been functioning for months now |
|
|
|
I've got the computer set up with the appropriate software and I've been running it from my basement without any problems for a couple of days now. The GUI tends to sag a bit, especially at full screen mode on my second monitor, but that's okay since I only need to monitor the miners, temps, etc. The only missing feature I can think of would be the ability to run the program from my Android smartphone. It would be awesome to be able to access remote desktop from anywhere.
The ambient temperature is almost 10 degrees lower down there as well, which is definitely one of the advantages here. I've thought about moving the PC into my garage or patio enclosure to keep it at an even lower ambient temperature, especially with winter coming (located in Cleveland, OH whose climate borders on frozen tundra for most of the year). I worry about condensation though if its left in either of those rooms because neither of them have humidity control. Is this something to be concerned about or is it a non-issue?
Thanks to everyone for your input. I'm glad I got this figured out because it's much quieter in here now.
Depending what remote desktop software you are using, most have some sort of android app available now. I haven't had to look that hard, I just use connectbot to SSH into my stuff at home. |
|
|
|
|
I switched from linuxcoin to bamt, but that was mostly because I wanted to setup multiple rigs quickly with very little config. My office with desktop is on the second floor, and my rigs are in the basement, so I mostly do exactly what you say, actually :p
|
|
|
|
Right, you asked, 970A-UD3 (£65) - Fried on the 24V connection and ruined my Akasa AK1200W PSU in the mean time MA770T-UD3 (£55) - Just died, completly P67A-D3-B3 (£90) - WOnt accept more than 3 VGA cards, even though it as 5 PCI-E (2 x16's, 3 X1) and when you add PCI-PCI-E converter it just doesnt boot full stop. Now, MSI, 870-C45 (£45) - Doesnt skip a beat, must be 2 months solid now no down time p67a-gd65 B3 (£90) - Pretty new but looks and works a million times better than any gigabyte i've had. Rock solid now. Gigabyte have cost me big time, That PSU was £145 on offer from £195 so all in all they cost me £300+ Nothing any of you can say to change my mind, MSI (as back in the day) kick ass  Well, two months isn't that long :p Also, for anecdotes, I had an MSI board many years ago, which had an onboard video card blow out, and then a replacement... which had the onboard CPU socket die, eat a CPU, and then proceed to eat another cpu I put in it trying to find out what went wrong. That is when I got mad and bought a gigabyte :p Fact is, over a large enough cross section, you will generally find someone with terrible experience with one brand or another. i think the warranty handling is more important... But then, I am lazy and tend to buy local so I can bring it back and make it their problem, so I don't really have to deal with the manufacturer much. On a side note, I stopped buying from new egg when my one gigabyte board did die and they wanted to bill me return shipping and a restocking fee to refund it.. sometime within 2-6 weeks due to shipping it from Canada (point of purchase) back to california. I will pay $4 extra on my $500 order next time to deal local, thanks :p |
|
|
|
Is there anyone who would be willing to look into what it would take to switch out the phoenix miner with cgminer in the bamt stack? Is this possible?
I am assuming the changes would need to include:
- Startup cgminer instead of phoenix at startup.
- gpumon could be replace with the cgminer interface or gpumon could read the output of cgminer.
- status reporting for mgpumon would need to read the statuses from cgminer.
- anything else i am unaware of.
I would be willing to pay a fixed price (BTC or $$$) to see this completed and working with the existing bamt toolset. It would also be nice if this could be a patch that can be run via /opt/bamt/fixer
Yeah, this was discussed earlier, cg miner doesn't really output a lot of the data needed: You would need to make changes to cgminer as well, I assume. |
|
|
|
Yup, yet another cheap and nasty component off a gigabyte board as killed my rig. It's a 5V mofet at best guess, everything turns on but no life and the CPU fan doesnt move (yet its fine).
Just steer clear of gigabyte, there cheap!
My five rigs have been running $84 gigabyte boards for several months no with no problems. My desktops have all been gigabyte boards for the past 5+ years, and I have had exactly one bad board, (recently), with a DOA lan port which I had refunded from new egg. Other then that, I moved to them because I was sick of other motherboards dying randomly and/or taking parts with them. |
|
|
|
|
yes, running a motherboard with an onboard gpu means it will consume the same power if you use it for a display: There are, however, two problems with this: a) why would you have bought a board with that to start with, and b) why would you be displaying something on it? I have six rigs with multiple cards, with uptimes measured in months, and see no appreciable difference in stats on primary/secondary/etc cards.
OTOH, 80+ efficiency is ignored too often, power supply efficiency is huge!
|
|
|
|
For some reason, one of my 5850 cards drops off and stops mining at random intervals between a couple of hours and a couple of days. It's not due to overclocking since it does it even at stock speeds. Not due to heat, temps are around 53C while mining. So here's my kludge to get around this and not have to check on the miner all the time: crontab -e add line: @hourly /etc/init.d/mine restart Save. Done. It restarts the mining process every hour. So far it works like a charm. that'll work, but it sounds like you've got a defective card. if under warranty i would return it. also, why isn't the automatic dead miner detection taking care of this for you? it does exactly the same thing as a mine restart, and does it every time you go X seconds without a share... is that not set up, or is it really not working? Yeah, the auto dead miner restart would work better, because this way (assuming you have multiple cards) They are all down for a few seconds every hour. It' not a huge impact, but it will cut into your total average hashrate. If it does it at normal clocks as well, and only the one card, you might want to consider applicable warranty. |
|
|
|
Jusstas a note - your payment timeplines are not realistic. Mining perations ar not run by businesses, you will find a problem with someone dwealing with this daily on a constant basic. People like taking a day off sometimes.
These payment timelines would be fairly clearly based on automation, or else I guess someone with way too much time on their hands. I don't have this sort of capacity, sadly. |
|
|
|
I try not to have a "manager node", in case that is a single point of failure. I use a lazy solution on all the rigs, this script runs via crontab: #!/bin/bash
URL="http://server/mining/temp.php?id="`hostname`
TEMPS=`export DISPLAY=:0;sudo aticonfig --odgt --adapter=all | grep Temperature| awk '{ print $5 }'| sed 's/^/\&temp[]=/g' | sed -e :a -e '$!N; s/\n//; ta'`
wget -O /dev/null $URL$TEMPS
On the server, temp.php $server = $_GET['id'];
$temp = $_GET['temp'];
$i = 0;
foreach ($temp as $t)
{
//write to log file/DB
}
The lazier yet answer is BAMT, I used to have logs running, going to a single node, etc, but now I just run mgpumon and everything is there :p |
|
|
|
I have an old eeePC netbook running dhcp and netboot for all my miners. The MAC address to hostname/static ip mapping in ISC dhcpd is a cakewalk, and has no limitation on the number of mappings you can make. The miners get their hostname from the dhcp server as well, so I can reference them by name or by ip.
The netbook pulls about 11W and has its own battery backup, and the miners don't need any local storage at all. A couple edits to the boot server and I can have a new miner up and running in minutes.
If your router for some reason will not let you assign static IP's, or for some reason caps out at 20 (I would have thought better of cisco, but I guess they want you to buy the next biggest one...), then a netbook, old pentium type machine, etc will make a fine router. There are a few linux distros just for this sort of thing, the one I used ages ago was smoothwall. |
|
|
|
Hi, I'm having some trouble setting up my rig and I was hoping some experts here can help. I am hoping someone can assist me on my setup through AOL instant messaging. I can pay in Bitcoins for whoever can help me get my rig running. I want to setup a rig with 3x ATI 6870's and 2x 6950's, the mobo I'm using has 4 pcie x16 slots and 1 pcie x8 slots. I'll be using a riser cable on 1 of the cards and a pciex1 to pciex16 riser cable on the other. If I can't get all 5 cards to run, I'm willing to settle for 4 cards running. OS is Windows XP. I've tried to do it myself but it is not recognizing all the cards. I've tried plugging a monitor in it and still nothing. Sometimes I can get the system to recognize maybe 2 cards but not every card I put in. I don't have dummy plugs but I have tried to plug another monitor in it still doesn't work. Please reply if anyone can help me. Also, how much Bitcoins do you charge (please give me a quote in USD, I know that bitcoin's price fluctuates and I would like to get a fixed amount in USD)? After setup I would like to let the rig run for 1 day before submitting payment, to make sure nothing goes wrong. (I have setup a rig before and the rig just shuts down and restarts itself). I really like to get by rig up and running because right now I just got a bunch of cards sitting around doing nothing  . Here are some of my specs Mobo: MSI 790FX-GD70 (it has 4 pciex 16 slots, 1 pcie x8 slots, and 1 pci slots) Gpus: 3x 6870, 2x 6950 PSU: 1200w watts Rams: 2 gigs HD: IDE 18 gigabyte CPu: don't exactly remember but its some slow AMD cpu that most rigs use. Thank you all for your time.  Give BAMT a try, it auto detects the video cards, no terminators or monitors needed, and you just need to edit two files to get it up and running: About 60 seconds of work. Instructions to install it to a USB stick are on the BAMT page: http://www.aaronwolfe.com/bamt/. |
|
|
|
Use reservations in the router so that the same MAC's get the same IP's every time. Easier to do that then assign statics to each machine. Basically the same thing but easier to implement.
This. And, since you set it by MAC address, if you have to wipe network settings, try a new OS, etc, they still get the same IP... and a good router means it is all a handy list all in one place. |
|
|
|
For anyone who's interested; I've forked the radeonvolt project and made some cosmetic changes to the code. Well, one functional change in that the program isn't restricted to the HD5850 anymore. It should accept all ATI cards and check to see if the correct VRM chip is in use, and try to proceed if it is. Also, the subvendor (XFX/ASUS/Sapphire etc.) is now displayed with the device information, and a --debug option has been added that prints out extra (more or less necessary) information. Github pageI'm still trying to find out if GPU-Z can read the VRM temperatures of my card, because if it is, it should be doable in Linux too. I have an XFX 5830 that it would be neat to be able to use radeonvolt on, and an asus 5870 (Which I expect less, as I know it is a super-special voltage regulator), and asus/HIS/HIS iceq-x 6870's. Let me know if there is any test info I can provide from them :p |
|
|
|
Let me know if any of you know of a good service that you use thank you! [/quote electronstorm.ca offers a lot of vps services, with many different options, and location in many datacenters. Disclaimer: See my sig :p |
|
|
|
I have seen a lot of cheap boards with PCI-E x 16 and PCI-E x 16 (4x4) any differences between the two? Does it mean I can put two GPUs?
For mining, a pcie X1 slot is fine, a GPU will not fit in it, but you can get cables to convert the slot for reasonably cheap. You don't need nych bandiwdth to the card at all. |
|
|
|
750 sounds good but without knowing the amperage distribution not very helpful, are these still equiped with the raid drives? One card horizontal is a given, two might be possible depending on the specifics of your units. Agreed with Deslok about the power. If it looks like the chenobro at: http://usa.chenbro.com/corporatesite/products_detail.php?sku=19Then I'd say (just from eyeballing it) you could likely get two video cards "floating" horizontally over the motherboard if you could rig up a bracket. I don't know about more, would depend on the card, and if it is in a rack or not (if you can cut a hole and vent heat out the side or not :p) |
|
|
|
Because of the oddities of your particular setup(pci->pcie adapter) we would best be able to help you if you had some pictures we could look at
In 2U, you will likely pretty much have to mount them horizontally. The pictures might help... are the pci-pcie extenders cables so you have some wiggle room? It also depends what else you have in the case already... |
|
|
|
Anyone ever heard of a datacentre using equipment other than air conditioning, I haven't.
For peace of mind I run my expensive hardware investment with an air conditioner that will always keep the room at 19 celcius - I can go to work, go on holidays, watch tv and be assured that my machines are at a steady temp.
My rigs use 5kW/h and my air con can cool 7kW of heat with just over 2kW of power. Why invest big and then skimp on cooling? Cooling needs to be factored into the cost and is as important as power surges or dirty power - it will destroy your hardware!
With air conditioning you can't introduce separate systems as it messes with the airflow. My GPU's on my 6990's run at around 70 celcius or under continuously. This is with 3x 6990's in each rig running at a standard 830Mhz per GPU, they are also running inside haf x cases with a modified side panel (professionally designed and cut via an engineering company) to fit 4x120mm delta fans for air flow. These fans have dust filters also. Each rig uses about 1.2kW/h and there are four in total. Just to add the fan speed set on the 6990's themselves is 70%, though it can manage no problem at all on 60%.
If your serious then professional is the way to go. Jobs isn't still running his company out of his garage. Business (i.e. investment) requires industry standards. Otherwise it's a joke.
I see people running custom cooling solutions and are just about keeping them under 85 celcius. What if it's a hot day and they pop up to 95? Do you panic? Do you call in sick to work? Do you fly back from your holidays?
I am also considering down clocking my GPU's rather than having them maxed out at 830Mhz all the time - a few less hashes isn't going to kill me. With regard to overclocking, I wouldn't consider it. Cards are certainly not designed to be maxed out all the time, never mind overclocking them.
google for "Data center liquid cooling". 3 million results. It isn't common, but it is not heard of. Keep in mind, bitcoin miners with 4 or even 8 GPUs per system is a lot of heat output per sq foot, well into or past the more exotic data centers like blade servers, etc. Also, on the note of 25kw of lighting, keep in mind that a high density server cabinet can hit 30 kw all by itself right now, with projections up to 50kw coming in a couple years (Source: http://www.42u.com/liquid-cooling-article.htm). Was all 25kw of lighting confined in a 42u sized space, and, also important, all stacked on top of each other? On that note, remember, that is one cabinet, you sure can get a lot of those in one room. Also, most data centers these days are looking into a lot of other options to improve PUE, heat wheels, liquid cooling, etc, at a certain point, it isn't efficient to just throw more AC at it. |
|
|
|
Air is a liquid
Air is free
Air is easy to move around
It also has a thermal conductivity that is 25x lower than water. so use 25x as much air...problem solved. At a certain point, the 25X airflow becomes a noise and possibly cost problem: I mean, sure, high CFM fans are cheap, but 25 is a dam big multiplier. If the profit margins on mining had stayed at the (clearly ridiculous, but a guy can dream) levels of $30/BTC, then it would almost be worth it to cool it with flourinert. However, 3M wants... a lot for that stuff. A lot a lot. You could make it nearly silent. There is at least one company I read about specializing in data centre cooling with flourinert, and the savings in cooling power is impressive. You have to remember, of course, that it depends on scale: What is easy to aircool in your house might not be easy (or possible) to aircool if you have, 5, 15, or 50 of them. |
|
|
|
|
