Topic: CGWatcher 1.4.0, a GUI/monitor for CGMiner & BFGMiner to help minimize downtime - page 5. (Read 402542 times)

Use your head. Have cgwatcher run at startup and set it to make sure the miner runs.. fuck its not hard. You got a brain, fucking use it. If cgwatcher doesn't run at startup on win8 use a BATch file to run cgwatcher. Don't know what a BATch file is, here: http://bit.ly/tHam0T

Yes I've got my crankypants on, simple shit...ffs

Hi guys,

Can someone explain how to set CGWatcher so that it resumes mining if PC restarts after a power drop? I have Win 8.

Nope, if you get a bsod or other type of windows crash you need someone to hit the reset/power button
With linux you can 99% of the time reboot remotely via ssh/putty.

A good solution is a remote wall outlet/pdu, equipped with a phone line or ethernet port.
Either that or making an arduino remote reboot device.

Is there an option to restart computer if it is freezing? Or any other possibilty?

Yeah. It's a common mining board, Gigabyte 990UD3. I was able to eliminate CGwatcher as the cause. It seems to be a Win8.1 issue. I don't think the board is to blame as I ran BAMT and it mapped perfectly - plus, the mapping is the same in both OS.

I run most of my rigs off of BAMT, but I'm toying with Win8 because of CGwatcher being able to keep it going. I considered doing a Linux script to parse the log files and restart the miners based on that.

It'll either be windows to blame or the motherboard. In any case I've seen this on 1 or 2 motherboards but it was never a critical or build-breaking issue. In one case it was because of the on-motherboard GPU intel processors' GPU being enabled but we couldn't disable it without causing other problems (was an Asrock Z87 board).

I would suggest trying BAMT or check out the mining support subforums as these problems can occur even without cgwatcher getting involved.

So I've got a weird issue. I've replicated it twice now, trying to narrow it completely down. Issue is a multi-card system gets corrupted so that the temperature/RPM sensors get mapped to the wrong cards. Anyone seen this? Weird thing is it survives CGwatcher, so CGminer will use the incorrect mapping when ran by itself, even after reboot. Only fix has been to reinstall the OS as trying to reinstall drivers results in a lockup. I'm guessing driver corruption.

- Using Windows 8.1 Enterprise
- Latest CGWatcher

Card | Temp Sensor Mapped To | RPM Mapped To | Hashrate Mapped To
GPU0 | GPU1 | GPU0 | GPU0
GPU1 | GPU0 | GPU1 | GPU1
GPU2 | GPU2 | GPU2 | GPU2

This will cause one card to go SICK/DEAD/OFF because it may see a low temp on the incorrect card and thus not adjust its fan.

Trying to narrow down how I'm doing it. But both times the system has worked fine without issue for a day or two. This seems to happen when I run it with CGWatcher.

*edit*

Further testing I did a system restore to before the issue and now GPU2 is using GPU1's temp sensor. Such a weird issue.


---------------

I did notice that someone else suggested the ability to run two profiles at once without using two installs of CGWatcher. I'd second this. I use the new "xIntensity" for a few of my cards, but some do not like it, so I have to run multiple miners for all cards to work correctly. Be a useful feature.

Setting CGWatcher's compatibility to Vista (Service Pack 2) seems to work. Weird

good jobs

Can you add to the monitoring a setting as follows:

If HW count is greater than 0 then the computer/miner is restarted.

I use sgminer and a experimental kernel and it seems to get HW errors after 20 hours or so of mining.

I'm using the restart miner after X hours of continual mining but it would be good if there is a setting to only restart if HW errors are greater than X.

Also can you change the monitoring setting "if total hashrate fall below" to be a drop down and you can specify if its hashrate or WU. As WU is what you are doing at the pool and not locally? Sometimes sgminer needs a restart as the kernel doesn't do as much work at the pool ever tho hashrate is quite high, a restart fixes this. So it would be good if cgwatcher could restart the miner if WU falls below a certain amount.

How do I keep this thing from modifying my GPU fan settings when it restarts bfgminer? I set my GPU settings externally with Sapphire Trixx, along with my fan speed, for a reason, and if it restarts bfgminer while I'm away, it turns fan control back over to the GPUs, and then I'm stuck hashing at 65% capacity until I get home.

Sounds like your card and/or mining parameters are not stable: not necessarily a problem with cgwatcher.

Maybe you should ask in the relevant sub-forum if it persists.

When my video driver crashes sometimes one video card will display a crazy hashrate. I know there's a reboot on "if under X hashrate" option, is there anyway to do a reboot "if over x hashrate"?

When I open CGWatcher I get a "CGWatcher has stopped working" error.. I've tried a restart, a previous version, and different file locations all to no avail. I'm running Windows 7 Home Premium 64-bit, and NET Framework 4.0 is installed. What could be the problem? CGWatcher says "Detecting OpenCL devices..." in the background of the crash.. not sure if that helps any

Another way to do this is to set the scheduler to switch profiles for a % of the day to mine for your dad.

I got a strange problem with cgw 1.3.5.2 ... every day it just says "not mining" and in log it says something about not sending shares, then it tries to kill cgminer process, then it can't start it again in 3 attempts, and it just stays there and do nothing. Of course if I click "Start" it normally runs cgminer and everything is good. In options I set to restart the computer if the thing happens, but cgw doesn't restart the rig. What could be the problem?

You can keep them in the batch file and just tell CGWatcher to use the batch file instead of pointing it to the miner's executable. In Profile Manager, for "Miner Path" set it to the batch file instead of cgminer.exe or bfgminer.exe.