[1050 TH] BitMinter.com [1% PPLNS,Pays TxFees +MergedMining,Stratum,GBT,vardiff]

[Turbor]

Sending over some hashes soon. You will be GPU maxed

[Beaflag VonRathburg]

You're welcome:

639 45244 2012-03-03 12:35 5h 32m 473,702 547,532 57.9% 102.064 NMC BeaflagVonRathburg confirmed
638 169472 2012-03-03 12:35 5h 32m 475,886 1,496,978 27.2% 102.514 BTC BeaflagVonRathburg confirmed

Beaflag, can you do it again please 

I've been having some serious stability issues. I'll get back on there if I can get it running again, but need some help.

- MSI H61-p21 MATX board
- G440 1.6ghz Celeron
- 1gb DDR3
- 4x 5870 on unpowered risers
- Cougar GX-1050 PSU
- 240v 20amp line

I had it running stable for a while, but I then bumper the clocks up to 960/170, from 950/170. I left it running and went to work. I noticed while at work that the machine went down for some reason. When, I got home I tried to restart it, but issues. I think it crashed X as it was still on, but not hashing. It will post, boot to the 11.04 menu screen, but when I try to start it just hangs at the Xubuntu 11.04 screen with the dots bellow it. Finally, it says:

unmount: can't unmount /cdrom: Device or resource busy.

My sata connectors are turned off in the bios so it doesn't make sense to me. I've tried clearing CMOS, but it didn't do anything. I know that this probably isn't the place for this, but I'd like to get this running soon.

[P4man]

I've been having some serious stability issues. I'll get back on there if I can get it running again, but need some help.

- MSI H61-p21 MATX board
- G440 1.6ghz Celeron
- 1gb DDR3
- 4x 5870 on unpowered risers
- Cougar GX-1050 PSU
- 240v 20amp line

This is probably not a good idea. No way that mATX motherboard was designed to deliver that much power to the PCIe slots.
Remove one or two cards, and see if it runs stable. If it does, you are going to need powered risers.

[jake262144]

This is probably not a good idea. No way that mATX motherboard was designed to deliver that much power to the PCIe slots.
Remove one or two cards, and see if it runs stable. If it does, you are going to need powered risers.

Roger that, the board has just one PCIe x16 slot and three PCIe x1 slots clearly not having been engineered for multi-GPU computing.
The fact of the matter is, only PCIe graphic cards are allowed to draw up to 75W per slot - all other PCIe devices are limited to 25W.
Do note that motherboards officially supporting triple- or even quadruple GPUs are equipped with additional power connectors for powering the cards up.

Fitting multiple PCIe x16 slots can be regarded as a kind of declaration by the manufacturer.
A typical mini-ATX board is what it is - an inexpensive piece of equipment meant for low-power PCs.

Did your combination of 5870s ever work in a stable fashion with this board?

Don't worry about that umount nonsense, the persistent liveusb version which you must be running is mounted as /cdrom.
While the system is running try issuing the command
cat /proc/mounts | grep cdrom
The result will point to your usb drive (/dev/sd*)

[WhitePhantom]

Did your combination of 5870s ever work in a stable fashion with this board?

I had it running stable for a while, but I then bumper the clocks up to 960/170, from 950/170.

That's the part that has me scratching my head.  Why'd it work stable until the overclock was increased and then fail to regain stability after returning to the original stable clocks?

Remove one or two cards, and see if it runs stable. If it does, you are going to need powered risers.

I would also be interested in knowing the results of trying it with only two or three cards.

[DeathAndTaxes]

It is right there in the quote you quoted.

"If it does you are going to need powered risers."

GPU can put 75W from the bus but most likey pulls no more than 35W.  However a mATX board likely wasn't designed for 120W+ from the bus.  If lucky you simply need powered extenders to take load off the bus.  If unlucky you damaged the VRMs on the motherboard.

[Beaflag VonRathburg]

This is probably not a good idea. No way that mATX motherboard was designed to deliver that much power to the PCIe slots.
Remove one or two cards, and see if it runs stable. If it does, you are going to need powered risers.

Roger that, the board has just one PCIe x16 slot and three PCIe x1 slots clearly not having been engineered for multi-GPU computing.
The fact of the matter is, only PCIe graphic cards are allowed to draw up to 75W per slot - all other PCIe devices are limited to 25W.
Do note that motherboards officially supporting triple- or even quadruple GPUs are equipped with additional power connectors for powering the cards up.

Fitting multiple PCIe x16 slots can be regarded as a kind of declaration by the manufacturer.
A typical mini-ATX board is what it is - an inexpensive piece of equipment meant for low-power PCs.

Did your combination of 5870s ever work in a stable fashion with this board?

Don't worry about that umount nonsense, the persistent liveusb version which you must be running is mounted as /cdrom.
While the system is running try issuing the command
cat /proc/mounts | grep cdrom
The result will point to your usb drive (/dev/sd*)

I'd like to thank everyone for their assistance. The machine ran fine for three days at 950/170 and averaged around 1650 MHS. I actually have two of these machines that contain the exact same parts. I managed to bring the second machine online on another pool for testing purposes and it has been running without fault for almost 3.5 days now with the cards in it at:

0: 975 / 170   v1.100
1: 850 / 1200 v1.163
2: 975 / 170   v1.100
3: 975 / 170   v1.100
I have one weak card in that system that for some odd reason will not overclock, but that's another issue. I don't

I restarted the machine in question twice to disable some items in the bios, lower the ram voltage, and change some options. I changed the clocks to 960 around 12:30am and around 11am I noticed that it had gone down. I can check the pool's time line of my shift work to see the exact time, but I don't know if it would help any.

I've been working on it now for about an hour attempting to trouble shoot and here is what I have found. I am currently using the machine from the live boot CD to post this. I had the USB flash drive that I have the OS on installed. It would only sit at the Xubuntu 11.04 screen and eventually after about 30-45 minutes it would display the mounting error message. I booted the machine from the live CD with the flash drive still in the back by changing the boot order in the bios. With the flash drive in, but booting from the CD I had all sorts of odd issues. The flash drive did not immediately mount, but rather I had to remove it, and reinstall it for the drive to be seen. With it mounted, I went to explore the drive, but every time I attempt select the drive it shows for a split second and closes. I cannot access the files on the drive through a visual source. I haven't tried to terminal run anything as I'm pretty much a Linux noob. The machine would also not allow me to open any program or connect to my internet source. I removed the drive and am now running from the boot drive without any issues. I'm wondering if my physical reset of the computer via the pins somehow screwed up my flash drive. I'd really like to avoid having to reload the OS to a drive and reinstall all of the software, but that is my next logical thought process. As much as I would like to avoid it I think I'm going to give creating a new liveUSB from another flash drive a try unless anyone has any other thoughts?

[jake262144]

I haven't tried to terminal run anything as I'm pretty much a Linux noob.

Very well then, try this:
(1) with the USB stick inserted, boot up the system
(2) when you reach Xubuntu desktop, press the [Alt]+[F2] combination
(3) type xterm in the window that pops up
(4) a text-mode window will launch, that's the terminal
(5) issue the command sudo cat /proc/mounts | grep cdrom
     You should see a line beginning with something like

Code:

/dev/sda1 /cdrom

, this will tell you which device is mounted as /cdrom

     fdisk -l will give you plenty of info about mounted devices and filesystems, feel free to upload the output if it seems arcane.

     If you want to see the contents of the device mounted as /cdrom, type

Code:

ls -lh /cdrom

[P4man]

It also sounds as if your stick has gone haywire. Try another stick.

I still stand by my point that you are overstressing your motherboard way beyond what is was designed for though; it may or may not work for some time, it may go up in smoke, it may give all kinds of weird problems. Its not worth the risk and time troubleshooting IMO, just get some powered risers. Even if it wouldnt solve your current problem, it will likely prevent other, and possibly more serious ones, like a smoking motherboard.

[jake262144]

...I still stand by my point that you are overstressing your motherboard way beyond what is was designed for though...

Preach to the choir, P4 
The fact that he has two such boards and one has been running ok might indicate that the PCIe GPU power load is dangerously close to the design limit, a combination of four highly overclocked 5870 might be an edge case here.
At least use powered adapters for two of those four cards, this way you'll only need to purchase four adapters for both rigs - not a huge expense by any stretch of imagination.

[boozer]

Doc,

Is this the worse string of luck you have ever seen?   Sheesh!      Every namecoin and BTC in the history is in the red cdf %, lol....   I keep waiting for the luck to swing

D&T,

Did you not like p2pool, or just come back to help us out? 

[DeathAndTaxes]

Did you not like p2pool, or just come back to help us out? 

Temporarily.  I found a "glitch" w/ BAMT 0.4 and p2pool.  Don't have the time to upgrade those rigs till this weekend the other rigs already upgraded are still on p2pool.

Looks like I have wonderful timing.

[DrHaribo]

Yes, a lot of bad luck lately. I sure hope there is a lucky streak just around the corner.

Let's bring down this monster block now!

[testconpastas2]

Hi everybody!.   Nice pool and client but a little bit "bad luckyed" since i came here . did I do this?

i had a question I would like anyone answer me. In the luck graph  , each time axis segment is a month?


Thank you in advance.


Antonio.

[DrHaribo]

i had a question I would like anyone answer me. In the luck graph  , each time axis segment is a month?

No, each block gets the same width in the graph. I should probably add timestamps on each one though.

[testconpastas2]

Thank you for the answer.   I was thinking it will be more "visual" if that graph has a fix time width  so you can see at a glance  last weeks or months' luck.   (some bad luck months will have 20 solved blocks and good lucked ones will have 40 for instance).

IMHO now  its more difficult to figure out how was last months luck.


 

[xkrikl]

I have similar feelings as Antonio has ... is that bad luck bacause of me joining the pool? ;-)
I've directed my hashing power to your great pool month ago ... approx. where the Rewards graph starts today ... and I see that the pool since then generated only 1700 bitcoins instead of expected 2467 which is 30% less ... that's pretty bad luck! :-(

I like the pool and the interface. It's great!
Hope for better luck for the next month! :-)

Anyway I would have 2 suggestions to consider ...
- I'd like to see longer history for statistics as an option (blocks, shifts, ...)
- It would be also nice if the luck and reward graphs could go further into the past (again probably as an option)

Pajka

[DrHaribo]

Phew, kjakman finally ended that monster block. I hope we get some easier ones now.

Glad you like the pool, guys. And bad luck is just something that happens. Some time back we had +30% luck for a while. Luck will turn again, sooner or later! Let's hope it's now.

IMHO now  its more difficult to figure out how was last months luck.

Yes, but it makes it easier to judge luck in general when every block gets the same size in the graph. If unlucky blocks get much more space, it will look like we have bad luck all the time. Maybe a new graph would be useful though that shows luck per time instead of per block.

- I'd like to see longer history for statistics as an option (blocks, shifts, ...)
- It would be also nice if the luck and reward graphs could go further into the past (again probably as an option)

Both are on my list. I will first add pagination for blocks, shifts and transaction history. Luck and reward graphs are lower priority.

[WhitePhantom]

Maybe a new graph would be useful though that shows luck per time instead of per block.

This graph at Pool-X (LiteCoin pool) is really nice.  I think you have to login to see it, but I think seeing it is worth making an account.
http://pool-x.eu/blocksAuth

In regard to functionality, it's actually pretty similar to the one at BitMinter, but I find it a lot easier to interpret.

Somebody told me a while back that the chart comes from http://www.highcharts.com

[xkrikl]

DrHaribo did you change anything on the server lately?
Or anybody do you know why this following happend?

Till yesterday always when we found BTC block we found also NMC block and in the list of blocks they were listed BTC and next row NMC both with the same timestamp (eg. blocks 169688/45504 - in our blocks list rows 648-649).
Since then there are two things:
1) some rows are missing from my blocks list: 657, 659-660, 662, 669, 671-687, 689-691, 693, 695-698, 701-702, 705-706, 712
2) BTC blocks and NMC blocks don't come anymore in pairs as earlier:

• when BTC block 170048 was found (row 661) NMC block 45862 is listed 1 minute earlier (row 658)
• when BTC block 170495 was found (row 688) NMC block 46281 is listed 5 minutes earlier (row 670)
• when BTC block 170638 was found (row 699) NMC block 46403 is listed 1 minute earlier (row 694)
• when BTC block 170759 was found (row 710) there is no NMC block listed mined by the same user near by
• when the last BTC block 170838 was found (row 713) there is NMC block 46572 found by the same user 5.5 hrs earlier (row 711)

Pajka