I couldn't get the /getid command to work with my telegram bot. Just in case anyone else is having that issue, this guide helped me get around that:

https://www.forsomedefinition.com/automation/creating-telegram-bot-notifications/

Wow. Thank you dbolivar!

The slow boot time has always been my only issue with nvOC. Especially when I've got a sick riser that I'm hunting for. I commented out the swap line in /etc/fstab and BAM! It was like a rocket ship booting up.

SO loving this OS now. Even more than before! Thank you again FullZero for this awesome OS.

For all the guys having stability issues, try moving to Claymore 9.5. I had issues a few days ago that I mentioned a few pages back. Same weirdness with rigs going to 0 hashes or rebooting continuously. 

Manually rolling my systems to Claymore 9.5 fixed that for me.

That's what I was saying. You've got to load the full complied version or the old binary won't recognize the newer kernel.

Regarding sgminer-gm, I originally tried just adding ethhash-new.cl to the kernel folder but it's not recognized unless you have a re-compiled version of the main binary.

This is what my working sgminer-gm shows as the version: sgminer 5.5.5-gm-a

It's across two different locations. I'm moving things over to the new version of claymore as a test. If that's not it, I'll try a different pool.

Looks like Claymore has jumped to v9.5 in the last day or two. I'm going to roll that out and see if helps.

Thanks Tytanik. I'm purposefully not automatically rebooting because it masks issues with the systems or would take a whole machine down when only one card was acting up.

Just to be clear, it's been all of my rigs across multiple levels of overclock. 16 systems in total. These are rigs that have been very stable in the past but have all decided to flip out at the same time. The symptom is that all cards in the system throw GPU errors all at once. I was just wondering if anyone else started to see the same issues or had noticed an uptick in reboots due to automatic reboots.

Outside of Simpleming OS systems, I've also had a Windows system and a raw ubuntu system with Nvidia cards also drop off. I'm really starting to think its an issue with Claymore. It's just odd that it started so long after his last release. Maybe its an issue with the latest DAG file.

On a side note, I think it's super critical to get a new version of sgminer-gm loaded that includes the ethhash-new.cl kernel. That way, we have a fallback from Claymore if folks are having issues with it or just want to avoid the stability issues that are associated with his dev fee process. Last time that I checked, v1095, that new Kernel wasn't included. PM me if you want a copy of the version I've complied and put on my systems.

Anyone else having watchdog issues where the systems drop to 0 hashes and need to be rebooted? Or having issues with rigs rebooting a bunch.

I purposefully left the auto reboot setting off Claymore to highlight rigs with hanging card issues and had my systems pretty stable. Starting about 5AM this morning, I've been having issues with machines at two different locations where the watchdog timers show all GPUs with OpenCL errors. Sometimes a simple reboot takes care of things. Other times the systems require a hard power cycle.

So far, I've rebooted every one of my systems(16 in total) at least once with several needing to be rebooted multiple times. These are systems that have been very stable for days and in some cases, weeks. Continuing to troubleshoot here but wondering if anyone else is seeing similar things happening.

ETH: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 0.000 Mh/s, GPU4 0.000 Mh/s, GPU5 0.000 Mh/s, GPU6 0.000 Mh/s
GPU0 t=42C fan=47%, GPU1 t=41C fan=47%, GPU2 t=45C fan=47%, GPU3 t=38C fan=47%, GPU4 t=43C fan=47%, GPU5 t=42C fan=47%, GPU6 t=4
0C fan=47%
WATCHDOG: GPU 0 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 0 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 1 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 1 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 2 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 2 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 3 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 3 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 4 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 4 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 5 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 5 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 6 hangs in OpenCL call, you need to restart miner
WATCHDOG: GPU 6 hangs in OpenCL call, you need to restart miner

Thank you for the updates Fullzero!

Quick question, the v15 notes state that openssh server is installed but disabled. How do we enable it?

With multiple jobs tugging at my time and a significant adversion to downtime, I just couldn't afford the extra time needed to continue tuning.  I had so many GPUs come in that I didn't have a test bench anymore! With 5 z87 boards maxed out at 7 cards each that were super stable, I was just going to see if I could find a few more. Thanks again, you've given me the added problem of having to find a few more gpus in order to max out my z170a boards!

Now, if I could only get risers to stop dying on me.  Between some junky new USB risers and the old ones I have from 2013, it seems like anytime I have to move or reboot a system, there's a riser issue. I've been meaning to say something for a bit but I wanted to give a shout out to the guys from Miningparts. I just bought a bunch of new risers from them and they're my go to when a gpu starts acting flakey.

One last thing for all the guys asking about power, I was trying to understand the difference between what the cards were using versus what I measured at the socket(efficiency) when I found this great resource for power supply sizing:

http://www.coolermaster.com/power-supply-calculator/

I've been buying 1200s for my 7 GPU systems but this is saying a 1k should be sufficient.  That jives with what I've been measuring at the wall on my maxed out rx470 systems so I'm trialing a couple evga G3 1k power supplies.  If that works out, it'll save me about $100 per new build.

Tytanick,

Of the two Soothaa mentioned, I only tried simpleminer-RX-v1093-4.11.rc8-16.60-enableexe0ANDips0.img. That one worked so I never tried the second one he mentioned.

Thanks to Soothaa continuing after I gave up. I can now run 7 570s on my 170a boards!

Anyhow, I've been carefully watching my power usage due to having to make sure I'm balancing the load of my machines across circuits. I'm not doing anything with the bios tuning to try and lower my power but I do set all them to power stage 4. With that, a 7 gpu rig runs under 1000w at the wall or just under 9 amps.

Soothaa,

I spent a good bit of the day futzing with that machine and I'm convinced that it's just not going to support 7 cards until MSI releases a new bios that's going to allow us to disable the SATA controller. I have several z98 boards from them that support 7 cards and in each case, the option that allows use of the 7th card is the ability to disable the SATA controller.

With 7 cards on the z170a, you can get 7 cards to show up in the OS and Claymore but as soon as a stress is put on the cards, you start to see weirdness with the OS like hangups or dropping network. I believe this is due to a resource constraint. Possibly interrupts or memory addresses. I went through the OS and ensured that everything that could be disabled was. That included: Audio, HPTimer, and the Serial port. Could be that the next BIOS from MSI fixes things. Only time will tell.

The highlight of the day had to be the new USB risers. Back in 2013/2014, those were crazy expensive and difficult to get. Now days, they're cheaper than the risers I bought years ago. These new ones popped right in and replaced the older risers AND the jumper wires I'd had to put in the PCIe slots to force the motherboard to see the cards.

Anyone experienced any errors like this?

[ 1084.587016] amdgpu 0000:07:00.0: GPU fault detected: 147 0x05708801
[ 1084.587016] amdgpu 0000:07:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x00E0983C
[ 1084.587016] amdgpu 0000:07:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x06088002

I've walked the cards and even went as far as replacing the OS but it keeps coming back. So far, I'm only seeing this on the z170a board with the 6 470s.

Well....

Spoke too soon. That 7th cards causes all sorts of stability issues. Random reboots and hangs. With the jumpers, I've got 6 cards installed and active but the 7th just doesn't seem to want to work. Maybe it's still a riser issue but. I'll try again when the USB ones arrive later today.

All right guys, I beat on it till it finally submitted to my will! I tried newer versions of Ubuntu and even went down the windows path before I came back to the physical layer being the culprit.

I'm using the older style risers like Soothaa but have the newer USB ones on order. This I believe was the source of all my ailments. Anyhow, here's what I did to get it working.

MSI Z170a Gaming M5
Bios Revision: 1.D0
Settings:
HD Audio - Disabled
PCI Versions - Auto
Integrated Graphics - IGD
4G Decoding - Enabled

This is where I'd get 4 or 5 cards active on 6 installed. Or, 4 cards active on 7 installed.

So, I started creating jumpers for each of my 1x PCIe slots. With 7 cards installed, I have jumpers on the 2nd and 3rd 1x PCIe slots and a 16x riser on my 2nd 16x PCIe slot. Lastly, I've got a newer style USB riser on the last 16x slot.

I was scraping the bottom of the barrel on risers to get this board running and it shows. Anyhow, it's now pumping away with 7 active GPUs and I'll replace all of them once the new risers arrive.

Hi Sootha! Any more luck with this board? I just picked one up and it's been giving me fits.

So far, the best luck I've had is with the D0 bios and these settings:
PCI ver: Auto
4g Decoding: enabled
Integrated Graphics Enabled
HD audio disabled

I've got 6 cards plugged in but only 5 show in claymore. I suspect that the 6th card isn't enabled due to bios issues or something else. Possibly it's requiring UEFI since there are a lot of folks reporting success under windows. I'm going to troubleshoot this a bit more this morning.

I'm running the latest build so Smos. V1095 and it appears that has the latest kernel and amd  driver versions already.

Anyone else running one of these z170a boards and been able to get them to see 6 or 8 cards in Smos?

Tytanik,

I've switched back to sgminer-gm version 5.5.5-gm-a that supports the ethash-new.cl kernel by pushing them manually to all of my systems. Can I provide you the updated version of Sgminer and the kernel so that your next release doesn't overwrite them or could you refresh the compile on your side please?

Thank you!

I'm also on ethermine and my last 24 hours looks like yours. The higher your hash rate the more noticeable the drop. My effective hash rate went from just under 1GH/s to 942MH/s. If you look at the stats for the top 10 miners on ethermine, you'll see that almost all of their charts show similar drops.