|
I get to differ. I have rigs using M.2 to PCIe Adapters and they work just fine.
I got:
1x Asus PRIME Z270-A w/ 9x 1070's all on risers, 2 of which use M.2 to PCIe adapters.
2x Asus PRIME Z270-A w/ 8x 1080Ti's all on risers, 1 of which use a M.2 to PCIe adapter.
1x MSI Z270-A PRO w/ 7x 1050Ti all on risers, 1 of which use a M.2 to PCIe adapter.
All have been rock solid for months.
The trick is to get decent M.2 adapters and be sure to center them in the M.2 slot.
|
|
|
|
thanks for low down. Loosk like you live in house? DO you have to have special agreement or plan with utility? I live in condo so solar is pretty much non-option.
I live in a private single family residence on a fairly rural lot. I did have to do a Net Metering agreement with the utility and they installed a meter that reads both forwards and backwards and transmits daily tallies for Delivered kWh (to me) and Received kWh (to them) and Net kWh. When I produce more power than I consume, they credit me the kWh at the exact same rate I pay them. My rates are constant 24/7/365. I know other utilities pay you less for the excess you send them, and also have variable rates depending on time of day and season. I'm glad I don't have to deal with any of that nonsense. |
|
|
|
I have 10 rigs atm consuming about 10kw 24/7. I have 400A service but plan to add a 2nd 400A service to my shop building and move everything there in the spring. I've got S9's inbound in February I hope. I keep pretty close tabs on my usage:  My solar arrays helps push energy back into the grid during sunny days, even now 10 days away from the shortest day of the year. My POCO hasn't "stormed in" yet. In fact, I have been talking to them about getting commercial rates and what all is involved. They know I'm mining. I live in the sticks in Virginia and my POCO is a CO-OP. They have been really great to deal with over the years, but I'm sure YMMV. |
|
|
|
|
999 Sol/S @ 6.02 Sol/W (166W) not too shabby either.
|
|
|
|
Replace the complete XMR part in 3main with this and test :
I tried Stubo's fix first and that took care of it. If you still want to be to test the new XMR section, let me know. |
|
|
|
I should have read up just a tick. There is another $running and the kill that is killing the plusCPU miner: reup=0
# running=$(ps ax | grep -i screen | grep miner | awk '"miner" {print $1}')
running=$(ps ax | grep -v cpuminer | grep -i screen | grep miner | awk '"miner" {print $1}')
if [ "$running" != "" ]
then
# target=$(ps ax | grep -i screen | grep miner | awk '"miner" {print $1}')
target=$(ps ax | grep -v cpuminer | grep -i screen | grep miner | awk '"miner" {print $1}')
kill $target
reup=1
fi
sleep 2
echo ""
echo "LAUNCHING: MINER"
echo " "
# running=$(ps ax | grep -i screen | grep miner | awk '"miner" {print $1}')
running=$(ps ax | grep -v cpuminer | grep -i screen | grep miner | awk '"miner" {print $1}')
if [ "$running" == "" ]
then
if [ $LOCALorREMOTE == "REMOTE" ]
I commented out the originals and put in modified ones below them. That did the trick! I'm back to REMOTE and plusCPU starts, followed by the GPU miner and both are running happily now. |
|
|
|
Can you try REMOTE again with the watchdog turned off? MINER_WATCHDOG="NO" # YES or NO # Monitors the rig and automatically corrects the detected problems. Highly recommended to use this!
This will help determine if the watchdog is somehow killing plusCPU. Switched back to REMOTE and set WATCHDOG = "NO". Here are the events upon a reboot: Last login: Mon Dec 11 14:09:26 2017 from 10.0.1.212
m1@miner06:~$ screen -ls
There is a screen on:
2044.plusCPU (12/11/2017 03:12:03 PM) (Detached)
1 Socket in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There is a screen on:
2044.plusCPU (12/11/2017 03:12:03 PM) (Detached)
1 Socket in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There is a screen on:
2044.plusCPU (12/11/2017 03:12:03 PM) (Detached)
1 Socket in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There are screens on:
2076.temp (12/11/2017 03:12:07 PM) (Detached)
2044.plusCPU (12/11/2017 03:12:03 PM) (Detached)
2 Sockets in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There is a screen on:
2076.temp (12/11/2017 03:12:07 PM) (Detached)
1 Socket in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There is a screen on:
2076.temp (12/11/2017 03:12:06 PM) (Detached)
1 Socket in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There are screens on:
2186.miner (12/11/2017 03:12:09 PM) (Detached)
2076.temp (12/11/2017 03:12:06 PM) (Detached)
2 Sockets in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There are screens on:
2186.miner (12/11/2017 03:12:10 PM) (Detached)
2076.temp (12/11/2017 03:12:07 PM) (Detached)
2 Sockets in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There are screens on:
2186.miner (12/11/2017 03:12:10 PM) (Detached)
2076.temp (12/11/2017 03:12:07 PM) (Detached)
2 Sockets in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There are screens on:
2186.miner (12/11/2017 03:12:10 PM) (Detached)
2076.temp (12/11/2017 03:12:07 PM) (Detached)
2 Sockets in /var/run/screen/S-m1.
m1@miner06:~$ screen -ls
There are screens on:
2186.miner (12/11/2017 03:12:09 PM) (Detached)
2076.temp (12/11/2017 03:12:06 PM) (Detached)
2 Sockets in /var/run/screen/S-m1.
m1@miner06:~$
So it appears that watchdog is NOT what kills the plusCPU miner. |
|
|
|
Ah yes. Sorry guys. My bad. I forgot to change those back to CUSTOM and $HOSTNAME after running the update script. So only outstanding issues that I have now are related to plusCPU; 1. Only works in LOCAL, not REMOTE 2. Password being passed is x and not MINER_PWD I have: 0miner "pool_address" : "$XMR_POOL:$XMR_PORT",
"wallet_address" : "$ADDR",
"pool_password" : "MINER_PWD",
"call_timeout" : 10,
"retry_time" : 10,
"giveup_limit" : 0,
1bash # XMR : if plusCPU is "YES" replace with your XMR info ##
XMR_WORKER=$WORKERNAME
XMR_ADDRESS="crazydane"
XMR_POOL="us-east.cryptonight-hub.miningpoolhub.com"
XMR_PORT="20580"
3main if [ $plusCPU == "YES" ] && [ $AUTO_START_MINER == "YES" ]
then
HCD='/home/m1/cpuOPT/cpuminer'
XMRADDR="$XMR_ADDRESS.$XMR_WORKER"
echo ""
echo ""
echo "LAUNCHING: plusCPU"
if [[ `ps -ef |grep cpuminer |grep -v grep |wc -l` -eq 0 ]]
then
if [ $LOCALorREMOTE == "LOCAL" ]
then
guake -n $HCD -r plusCPU -e "$HCD -a cryptonight -o stratum+tcp://$XMR_POOL:$XMR_PORT -u $XMRADDR -p MINER_PWD -t $threadCOUNT"
else
screen -dmS plusCPU $HCD -a cryptonight -o stratum+tcp://$XMR_POOL:$XMR_PORT -u $XMRADDR -p MINER_PWD -t $threadCOUNT
fi
echo ""
echo "plusCPU process in guake terminal Tab (f12)"
echo ""
running=""
fi
|
|
|
|
Change from : guake -n $HCD -r plusCPU -e "$HCD -a cryptonight -o stratum+tcp://$XMR_POOL:$XMR_PORT -u $XMRADDR -p x -t $threadCOUNT"
else
screen -dmS plusCPU $HCD -a cryptonight -o stratum+tcp://$XMR_POOL:$XMR_PORT -u $XMRADDR -p x -t $threadCOUNT
To : guake -n $HCD -r plusCPU -e "$HCD -a cryptonight -o stratum+tcp://$XMR_POOL:$XMR_PORT -u $XMRADDR -p MINER_PWD -t $threadCOUNT"
else
screen -dmS plusCPU $HCD -a cryptonight -o stratum+tcp://$XMR_POOL:$XMR_PORT -u $XMRADDR -p MINER_PWD -t $threadCOUNT There is also some settings in 0miner, open 0miner in xmr section, change to "pool_password" : "MINER_PWD ",
Made the above changes except I used "MINER_PWD" instead of "MINER_PWD " and restarted, but I'm still getting: /home/m1/cpuOPT/cpuminer -a cryptonight -o stratum+tcp://us-east.cryptonight-hub.miningpoolhub.com:20580 -u crazydane.115 -p x -t 10
And: /home/m1/ASccminer/ccminer -a lyra2v2 -o stratum+tcp://hub.miningpoolhub.com:20507 -u crazydane.115 -p solar -i 21
So password is being passed for GPU miner, but not CPU miner. And for both miners, the last octet of the ip addresses is being passed instead of the hostname. |
|
|
|
|
I just edited my previous reply. It would appear that host name is not being passed either. Instead the last octet of the ip address is being used.
|
|
|
|
Does it quits on remote or local or both? Restart miner and try to catch the command by : or Copy the output and run it with from /home/m1/ .... to the end in guake terminal to catch if miner quits or some thing else kills it I don't have a local display attached, I run remote only. I captured the command that executes and when I execute it from a ssh session, the miner works fine. m1@miner06:~$ /home/m1/cpuOPT/cpuminer -a cryptonight -o stratum+tcp://us-east.cryptonight-hub.miningpoolhub.com:20580 -u crazydane.115 -p x -t 10
********** cpuminer-opt 3.6.3 ***********
A CPU miner with multi algo support and optimized for CPUs
with AES_NI and AVX extensions.
BTC donation address: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT
Forked from TPruvot's cpuminer-multi with credits
to Lucas Jones, elmad, palmd, djm34, pooler, ig0tik3d,
Wolf0, Jeff Garzik and Optiminer.
CPU: Intel(R) Core(TM) i7-6850K CPU @ 3.60GHz
CPU features: SSE2 AES AVX AVX2
SW built on May 18 2017 with GCC 5.4.0
SW features: SSE2
Algo features: SSE2 AES
Start mining with SSE2
[2017-12-11 12:45:25] Starting Stratum on stratum+tcp://us-east.cryptonight-hub.miningpoolhub.com:20580
[2017-12-11 12:45:25] 10 miner threads started, using 'cryptonight' algorithm.
[2017-12-11 12:45:26] Stratum difficulty set to 500054
[2017-12-11 12:45:29] CPU #4: 66 H, 28.78 H/s
[2017-12-11 12:45:30] CPU #5: 66 H, 28.43 H/s
[2017-12-11 12:45:31] CPU #0: 66 H, 19.10 H/s
[2017-12-11 12:45:31] CPU #1: 66 H, 18.81 H/s
[2017-12-11 12:45:31] CPU #2: 66 H, 18.74 H/s
[2017-12-11 12:45:31] CPU #3: 66 H, 18.53 H/s
[2017-12-11 12:45:31] CPU #7: 66 H, 18.11 H/s
[2017-12-11 12:45:31] CPU #8: 66 H, 17.94 H/s
[2017-12-11 12:45:31] CPU #9: 66 H, 17.73 H/s
[2017-12-11 12:45:31] CPU #6: 66 H, 15.76 H/s
[2017-12-11 12:45:44] CPU #8: 250 H, 18.93 H/s
[2017-12-11 12:45:44] CPU #0: 267 H, 19.87 H/s
[2017-12-11 12:45:44] CPU #3: 262 H, 19.65 H/s
[2017-12-11 12:45:44] CPU #1: 264 H, 19.71 H/s
[2017-12-11 12:45:44] CPU #4: 447 H, 30.59 H/s
[2017-12-11 12:45:44] CPU #9: 250 H, 18.96 H/s
[2017-12-11 12:45:44] CPU #5: 441 H, 30.21 H/s
[2017-12-11 12:45:44] CPU #2: 263 H, 19.62 H/s
[2017-12-11 12:45:44] CPU #6: 208 H, 16.32 H/s
[2017-12-11 12:45:44] CPU #7: 252 H, 18.97 H/s
[2017-12-11 12:46:39] Stratum difficulty set to 350009
[2017-12-11 12:46:39] CPU #8: 1046 H, 18.90 H/s
[2017-12-11 12:46:39] CPU #2: 1087 H, 19.65 H/s
[2017-12-11 12:46:39] CPU #7: 1048 H, 18.94 H/s
[2017-12-11 12:46:39] CPU #0: 1099 H, 19.85 H/s
[2017-12-11 12:46:39] CPU #5: 1673 H, 30.23 H/s
[2017-12-11 12:46:39] CPU #4: 1693 H, 30.58 H/s
[2017-12-11 12:46:39] CPU #9: 1046 H, 18.89 H/s
[2017-12-11 12:46:39] CPU #3: 1088 H, 19.65 H/s
[2017-12-11 12:46:39] CPU #1: 1090 H, 19.68 H/s
[2017-12-11 12:46:40] CPU #6: 905 H, 16.34 H/s
[2017-12-11 12:47:35] Stratum difficulty set to 245006
[2017-12-11 12:47:35] CPU #1: 1097 H, 19.75 H/s
[2017-12-11 12:47:35] CPU #7: 1055 H, 18.98 H/s
[2017-12-11 12:47:35] CPU #0: 1108 H, 19.94 H/s
[2017-12-11 12:47:35] CPU #9: 1054 H, 18.96 H/s
[2017-12-11 12:47:35] CPU #2: 1097 H, 19.73 H/s
[2017-12-11 12:47:35] CPU #4: 1702 H, 30.62 H/s
[2017-12-11 12:47:35] CPU #5: 1685 H, 30.31 H/s
[2017-12-11 12:47:35] CPU #6: 903 H, 16.26 H/s
[2017-12-11 12:47:35] CPU #3: 1097 H, 19.73 H/s
[2017-12-11 12:47:35] CPU #8: 1055 H, 18.97 H/s
[2017-12-11 12:48:31] Stratum difficulty set to 171503
[2017-12-11 12:48:31] CPU #8: 1055 H, 19.01 H/s
[2017-12-11 12:48:31] CPU #0: 1109 H, 19.97 H/s
[2017-12-11 12:48:31] CPU #2: 1097 H, 19.76 H/s
[2017-12-11 12:48:31] CPU #6: 906 H, 16.32 H/s
[2017-12-11 12:48:31] CPU #9: 1056 H, 19.02 H/s
[2017-12-11 12:48:31] CPU #4: 1702 H, 30.65 H/s
[2017-12-11 12:48:31] CPU #5: 1685 H, 30.34 H/s
[2017-12-11 12:48:31] CPU #7: 1057 H, 19.03 H/s
[2017-12-11 12:48:31] CPU #3: 1097 H, 19.75 H/s
[2017-12-11 12:48:31] CPU #1: 1099 H, 19.78 H/s
[2017-12-11 12:48:46] CPU #2: 296 H, 19.58 H/s
[2017-12-11 12:48:46] CPU #9: 284 H, 18.79 H/s
[2017-12-11 12:48:46] CPU #8: 285 H, 18.81 H/s
[2017-12-11 12:48:46] CPU #4: 462 H, 30.52 H/s
[2017-12-11 12:48:46] CPU #5: 457 H, 30.20 H/s
[2017-12-11 12:48:46] CPU #3: 296 H, 19.57 H/s
[2017-12-11 12:48:46] CPU #0: 301 H, 19.84 H/s
[2017-12-11 12:48:46] CPU #1: 297 H, 19.63 H/s
[2017-12-11 12:48:46] CPU #7: 286 H, 18.89 H/s
[2017-12-11 12:48:46] CPU #6: 246 H, 16.22 H/s
When I do a control-c to kill it, watchdog never attempts to restart it. I also never get email notifications that it quit. So for now, I'm just running that launch command from the ssh window and things are running smoothly again like on 19-1.4 EDIT: That launch command is showing -u crazydate.115 -p x where it should be -u crazydane.miner06 -p solar I would think. mph ignores the password, but is should be passing along the MINER_PWD set in 1bash should it not? In my case, that would be: # For zpool use MINER_PWD="$WORKERNAME,c=btc"
MINER_PWD="solar" # Set the miner password. Default: x
It just so happens that the ip address I have mapped to miner06 is 10.0.1.115. So it looks like instead of the host name, it is passing the last part of the ip address. EDIT2: I switched to LOCAL and that allows both miners to run. I get this: m1@miner06:~$ ps -ef |grep miner
avahi 905 1 0 13:09 ? 00:00:00 avahi-daemon: running [miner06.local]
m1 2067 2050 99 13:09 pts/20 00:08:16 /home/m1/cpuOPT/cpuminer -a cryptonight -o stratum+tcp://us-east.cryptonight-hub.miningpoolhub.com:20580 -u crazydane.115 -p x -t 10
m1 2295 1241 0 13:09 ? 00:00:00 SCREEN -dmSL miner /home/m1/ASccminer/ccminer -a lyra2v2 -o stratum+tcp://hub.miningpoolhub.com:20507 -u crazydane.115 -p solar -i 21
m1 2296 2295 0 13:09 pts/24 00:00:00 /home/m1/ASccminer/ccminer -a lyra2v2 -o stratum+tcp://hub.miningpoolhub.com:20507 -u crazydane.115 -p solar -i 21
m1 2306 1953 0 13:09 pts/17 00:00:00 screen -r miner
From the above, we have: -u crazydane.115 -p x -u crazydane.115 -p solar And it should be: -u crazydane.miner06 -p solar -u crazydane.miner06 -p solar |
|
|
|
Thanks Phil. Just ordered 3 of them. Also, I did get 2 more 1060's in today. Temps right now are:  I have the target temp set for 69 degrees. So the GPU fans will spin up as needed to maintain that. |
|
|
|
Either open 3main and edit the 2 lines and add :$XMR_PORT in this section : if [ $plusCPU == "YES" ] && [ $AUTO_START_MINER == "YES" ]
guake -n $HCD -r plusCPU -e "$HCD -a cryptonight -o stratum+tcp://$XMR_POOL:$XMR_PORT -u $XMRADDR -p x -t $threadCOUNT"
else
screen -dmS plusCPU $HCD -a cryptonight -o stratum+tcp://$XMR_POOL:$XMR_PORT -u $XMRADDR -p x -t $threadCOUNT Or run the update script again to fix it for you I edited those 2 lines in 3main and changed 1bash back to the new way with port on a separate line. The XMR miner now connects to the pool. However, the miner still exits within 20 seconds of starting for some reason. The screen window closes immediately when this happens. |
|
|
|
I have 19-2 running on my test rig that is also running plusCPU.
On a clean reboot, plusCPU starts up right away as usual, but as soon as the main GPU miner comes online, something (wdog?) kills plusCPU.
Is the pool address correct ? No need to add stratum+tcp:// to pool address any more Looks like the issue is that the port is not being passed to the miner, so it defaults to 80:  And this is what I have in 1bash: # XMR : if plusCPU is "YES" replace with your XMR info ##
XMR_WORKER=$WORKERNAME
XMR_ADDRESS="crazydane"
XMR_POOL="us-east.cryptonight-hub.miningpoolhub.com"
XMR_PORT="20580"
I fixed it for now by reverting back to: XMR_POOL="us-east.cryptonight-hub.miningpoolhub.com:20580" EDIT: Something is still killing plusCPU miner after a couple of minutes, even with the above change. I see it start up successfully but then it dies. |
|
|
|
|
I have 19-2 running on my test rig that is also running plusCPU.
On a clean reboot, plusCPU starts up right away as usual, but as soon as the main GPU miner comes online, something (wdog?) kills plusCPU.
|
|
|
|
It's a nice way to straighten up the cards... but the fans don't do much for airflow, as they're placed just in front of the GPUs' exhaust fans. Ideally those 120mm fans should be placed behind the GPUs to help blowing cool air onto the backplates, I reckon...
Another option I thought about was to reverse the fans, and add an angle piece at the bottom as a seal, and then drop a piece of Plexiglas across the top of the GPU's to force all air to be pulled in from the back. This way air would be sucked in from behind and pulled through the gaps between the cards and out the front. Once I get the missing 3 GPU's I'll record temps with the current setup, and then try reversing and adding the "lid" and lower seal. Might also try sealing the front sides as some air will be pulled in there as well. |
|
|
|
That's a very good idea and a cheap solution too - to ensure that cards are lined up standing correctly and with 3 fans mounted like that.
Whats the dimension like -- I will do some experiments this weekend. Thanks for idea.
The aluminum angle stock is 1/16" thick by 1/2". 3x 120mm fans measure 360mm across (duh) (about 14.25") so that would be the length to cut the aluminum stock at. I cut mine a little short since I was only going by what was needed to support the GPU brackets. To precisely mark the holes for the GPU brackets, I held the aluminum stock in place underneath the GPU bracket lips, and then used a permanent market to mark for the holes through the oval holes in the GPU bracket lips. I did a similar thing for locating the holes for the fans. Btw, I never did shut down the rig while doing any of this. Did want to loose the hashes, lol. |
|
|
|
Cool, very tidy. I look forward to hearing the temps on a fully stocked rig. I'd like to go this way from risers too.
Does the board connect to psu through all the pcie on the right (of the first photo) and then the connectors on the back are 'jumpers' to the cards?
Right the GPU0 is at 68 degrees with fan at 50%. Room is rather warm at 26 degrees. The other GPU's are in the mid 50's. The fans I used are pretty weak, so I might get more powerful ones if needed. One thing I thought about was to add another 1/2" angle stock at the bottom to "seal", forcing all the air to be pushed through the cards. I'm assuming the colorful board is running the fans at 100% speed but I'm not sure. Next time I'm in the BIOS I'll have a look to see if it tells me. (the 3 fans are connected to fan headers on the mobo) Yes, the PSU is feeding the colorful via 8x PCIe power cables and the colorful is, in turn, feeding the GPU's. |
|
|
|
I stopped by Lowe's today and picked up some 1/16" x 1/2" x 3ft aluminum angle stock. They were $2.79 each. I then cut it to length for the 8 GPU slots on the Colorful mining board. Next I drilled and tapped (6-32) for the 8 screws to tie it to the GPU brackets. I also drilled 6 holes to be able to attached 3 120mm fans. For about 30 minutes work and less than $3 in parts, I think it turned out pretty nice.     Once I tightened the screws to the GPU brackets all the cards "stood up" like they should instead of leaning to the right which they otherwise always do with no support. I plan to do the same to my Onda rig. Next up will be getting a $6 laminated shelf to mount the motherboard and PSU on to make the whole unit mobile. The remaining 3 1060 3G EVGA's should be here Tuesday from newegg's special they ran the other day ($189). I also have a 1200W Server PSU from parallelminer on the way. That 2400W Delta is way overkill for a 8x 1060 rig, but I had a couple of extra ones laying around so I used one of those for testing. |
|
|
|
|
Nope, both the Onda D8P and the Colorful boards are rock solid. Only change I made in the BIOS was to change to PEG for the video. Haven't bothered updating either BIOS either since they just work, but I probably will at some point. Running nvOS on them.
|
|
|
|
|
Show Posts
