Topic: CGWatcher 1.4.0, a GUI/monitor for CGMiner & BFGMiner to help minimize downtime - page 13. (Read 402530 times)

Share difficulty allows devices with higher hashrates to get work with a higher difficulty instead of a lot of work with low difficulty. For example, let's say a GPU gets diff1 work (the lowest difficulty). It hashes and gets one diff1 accepted share every minute. Now we buy an ASIC that mines 100x faster than the GPU. Instead of sending the ASIC 100x the work, we can just send it diff100 work that it has to solve to get one diff100 accepted share. This saves a lot of overhead. Pools pay in diff1 equivalent, so 100 diff1 shares should be paid the same as one diff100 share.




I answered this in the post you created. I would suggest trying a different pool because I believe the issue is coinotron.




What are you doing (or what is it doing) when this happens? What do you mean by "getting data"? If after it comes back you can create a debug report and send it to me, I'd really appreciate it. That will hopefully show in the log what happened. You can create a debug report by going to the Tests tab and clicking 'Create Debug Report' button, and my email address is in the readme.




I understand that may be unexpected behavior, but it is doing what it says it will do. If you stop the miner using the Stop or Pause Mining button, this option should not start the miner. It only launches the miner when it closes unexpectedly, or when CGWatcher starts - which is by design and was asked for by users.



I'm working on coin stuff for CGRemote right now, and it will probably all make it's way into CGWatcher as well. It's not as simple as combining the data - I either need to give the user an option to select which provider to use for each field, or create scheduling options for both data providers. I'm going with the first one... I just need time to code it.




If you set a scheduled action to trigger when the most profitable coin changes, it will do just that - with no regard for which coins you're mining. The action is triggered, and if set to switch to the most profitable, for example, it will look at all of your profiles and switch to the most profitable.

If you set a scheduled action to trigger when the most profitable profile changes, it will only be triggered when the most profitable coin that you're mining (and is linked to a profile) changes.

I'm not sure if I answered your question. If not, please try to explain a little more and I'll give it another shot. A list of your profiles and their coins would be helpful.

I second the problem. I'm using the program to mine SHA256 coins. I have the schedule setup to switch profile when the most profitable coin changes with no luck. My guess is that it ignores the coins on the profile and it simply goes after the most (scrypt) profitable coin. The scheduler is running and if I setup the scheduler to restart it, it does restart, just to come back to BTC instead of one of the other SHA256 coins

EDIT: somehow it started to work :-)

Any news on combining CoinWarz and CoinChoose data using the method I suggested? I'd really like to see that happen. Alternatively, a way to manually input the data using some sort of network API would also be fine, I could write the software for fetching and entering the data myself. I'll even share it, if there's interest.

I get not responding a lot in 1.3.4.x..

This seems to occur when its getting data (i/o lock?)? It comes back after a short while but older versions didn't do this. Also it seems to occasionally "wait" to query the API and gets a "not responding" after a few minutes it comes good..

Other than that nice job.

I just startecd using CGWatcher with cgminer, I've been using cgminer for months and everything was fine.
 Today, I started  CGWatcher, and I've kept running for roughly 30minutes, although cgminer says it's connected, and it's running, coinotron says cgminer isn't running (or it showing 0khs).

What am I doing wrong?
The only thing I've done differently is I added the following arguments:
-o stratum+tcp://coinotron.com:3334 -u [username] -p [password] -I 18 -g 1 -w 256 --thread-concurrency 15232

Are you using bfgminer? And are you mining a scrypt coin? If so, bfgminer does not report a valid number for total miner diff1 shares. I'll be changing the way this is reported in the next version.

A diff4 share is equivalent to 4 diff1 shares. Your pool gives you work of a certain difficulty, which you must meet or exceed to have the share accepted. The higher your hashrate, usually the higher difficulty shares you get. But you get paid in diff1 equivalent shares, so a diff4 share should earn the same as four diff1 shares.




I would suggest using System Configuration (Windows key + R, type in msconfig and press Enter). Go to Startup tab, uncheck the old CGWatcher startup entry. The System Configuration shows combined startup entries (registry + startup folder).

If you want to remove the entry, I'd recommend using Autoruns, which allows you to disable or delete startup entries. It provides a comprehensive list of all of the places an application can set itself to start automatically.

In the registry, CGWatcher creates an entry in HKCU\Software\Microsoft\Windows\CurrentVersion\Run for the current user, or HKLM\Software\Microsoft\Windows\CurrentVersion\Run for all users. You'll probably need to run CGWatcher as Administrator to save or remove the HKLM entry. Previous versions of CGWatcher would try to save to HKLM, and it would appear to work when in reality Windows actually creates a virtual registry and sticks it in some WoW3264Node key (I can't remember where exactly). If you're familiar with the registry and regedit, you could do a Find for CGWatcher.

Where does CGwatcher add an entry to start on Windows start?
I seem to have 2 CGwatcher that tries to start at boot and can't find how to disable the older

I'll make a note to fix this in the next update. I don't have the code in front of me and I thought I had handled any culture settings, but apparently I missed something. This is why I added the setting to disable formatting, just in case. At least that works. One out of two ain't bad.

Yes, disabling formatting, and I get values now for that column. So, I guess it doesn't like Swedish separators....

Apparently, I have this:

Culture decimal separator................... (,)
Culture group separator...................... ( )

Are you using the latest version of CGWatcher (1.3.4.4)? You can turn off formatting of coin data values in Coin Manager. This is the only field formatted as currency, so I'll have to see if there is something causing this. What is you system's culture settings?... specifically, what characters do you use for group and decimal separators? (eg. en-us default group separator is "," and decimal is ".") This can also be found in CGWatcher's Report tab and debug report. This field should be formatted using your system settings.

Well, the thing is that CGWatchers Coin Profability list show all zeros in the "Best Price" column. On http://coinchoose.com/ there are actual values there, so something wrong with the parsing of the data? Actually, bitcoin is the only one with a value, and that shows 0.00000001...

Sorry about the delayed response, I'll hopefully answer everyone's questions here:


Dekker3D: You should look into setting variables on your miners, so they can all use the same profiles and/or config files, but each can be customized. More information on variables can be found here: http://manotechnology.blogspot.com/2013/08/web-config-files.html. Let me know if you have any questions.


mb0erman: It looks like you're setting the same values for all three GPUs, but one of them doesn't like these settings. It may be the voltage as Tigggger said. You can set individual values for each card by separating them with commas. Example setting three GPU clock speeds to three different values:



burgergold: It looks like Tigggger answered this one too, but you would put those in a batch file. Then in Profile Manager, instead of selecting the executable (.exe) miner file for "Miner Path", you select the batch file.


bouncemeister, Chillance: Tigggger answered this too. There is a new setting in the latest version of CGWatcher to kill the miner when restarting it, as opposed to shutting it down (which restores clock speeds and voltages to original values). This means the voltage should stay the same through the restart. Other than the ability for the user to change voltage manually in the Devices tab, CGWatcher does not make any changes to voltage.


Uniphase21: You should not need dummy plugs. CGMiner and CGWatcher should detect all GPUs, whether they have a monitor plugged into them or not. However, Windows and CCC may only show GPUs that monitors are plugged into. Fortunately, CGMiner and CGWatcher don't care what Windows thinks.


Chillance: The "best price" is provided by your selected coin data provider (coinchoose or coinwarz). If the data in CGWatcher consistently does not match the data on your provider's website, please let me know. Because this information changes often, they may occasionally not match. But if you refresh coin data at the same time you check the website, they should show the same values.


Tigggger: Thanks for the help.

I run a rig with 3, 7950, all the same type, and CGWatcher running cgminer. I disconnect the monitor from the card in the first slot, after turning the monitor power off, and unplugging the monitor from the wall socket, and the rig remains running with no errors.

Win 7, 64 bit, ASUS extreme6 MB, intel dual celeron. Your mileage may vary according to the type of GPU you are using, and the type of motherboard and its BIOS revision.

Also, anyone here experience only zeros 0.00000 in Coin Profability -> Best Price column?

And, can someone be kind and provide the math to calculate price per day for my hash? Basically, I would like to know how I calculate what is shown over at coinwarz.com in "the Revenue /Profit (per day)" field using coinchoose.com.

Thanks!

My guess is you still need dummyplugs.

Will this take the place of dummy plugs in your mining rig or should I still put those in there?

Thanks tigggger, i'll give that a try!

In the meantime i hope voltage monitoring will be considered by the dev team. 

Despite having several identical rigs, this problem kept happening to me on one of them but not the others.

I made 2 changes and the problem hasn't occured again, don't think the 1st had anything to do with it (but including it just in case) almost certain the 2nd is what sorted it.

1. Changed from Sapphire Trixx to MSI Afterburner, used that to set clocks/voltages and selected option to apply clocks on startup
2. In the latest version of CGWatcher, use the two options to kill cgminer rather than quit, it says in the release notes that this preserves the gpu settings.

I agree with bouncemeister. Would be nice with a way to monitor voltage changes too. Because, as I said, if this changes to 1.2 from 1.050 for me for some reason, the temperature increases like crazy and feels rather unsafe.

good info
thanks bro