Topic: [Awesome Miner] - Powerful Windows GUI to manage and monitor up to 200000 miners - page 206. (Read 703563 times)

----------------------------
Troubleshooting
Use the GPU settings dialog, via the toolbar Tools -> GPU Settings) for adjusting GPU settings on a single GPU to verify that all settings works as expected
To verify if the GPU settings were applied successfully, use the Reload button in the GPU settings dialog to reload all properties from the AMD and nVidia GPU drivers.
If the values show up as expected, the GPU settings and overclocking operations have been successfully applied
If one or more settings didn't have the expected value, please report which GPU model and which setting that didn't work together with the Awesome Miner log file.
Do not use the MSI Afterburner application to check if the GPU settings were applied successfully. MSI Afterburner uses different overclocking concepts than other applications and may not display the current GPU settings correctly when set from other applications.
-----------------------------------


I have done a test and I do not see results. On one side of his help he says that you have to reload or update to see if the changes are reflected. I could leave a capture of that menu, sometimes the translations are not perfect.

It is not advisable to have MSI aftherburner installed, but it also does not indicate whether a change has been made. That's why it's important that you identify that GPU window for the native OC

If the native mode is not compatible with the RTX fans. How do we control them? Do they only go by car?

For me the important thing is to know if the OC changes are applied using the native mode, but I do not identify which part of the program I have to give to see the changes, I would appreciate a capture.



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

Bminer problems.

Now I have been seeing that if it is true that it does not need to include the -al-go in the miner's configuration. In fact it works for me with many algos, but when I tried it it was with tensory and there if it is true that it does not work, I suppose it will be the miner's own fault.

***
Did you tried Clocking Profile Groups like patrike mentioned? I think (might be wrong) you'd need that even if you separate the miners.

Hello Patrike, and thanks in advance for your attention.
I understand what you tell me and this is how I have done it. I explain it better.

I have two GPUs, one 1070 and another 2070.
I have 2 miners. 1070 and 2070
I have two profit profiles 1070 and 2070
I create an OC profile and apply it only to some algorithms of the profit profile 1070.

What I am saying is that when the OC is activated in profile 1070, it also applies in profit profile 2070 (and in this case it is clean of OC in the AM). When the OC is activated, the OC is applied to the two profit profiles and therefore to both cards and to both miners, when it should not be so, it should be applied only in the profit profile 1070, right?

At least in my case, so I do not know if I'm missing something in the configurations.

Regarding the second theme, yes, the curve I am referring to is the one in the "Voltage / Frequency curve editor"

Regards


EDIT:

https://i.imgur.com/kmnyksS.jpg
https://i.imgur.com/0eYN2V4.jpg

https://i.imgur.com/767C39S.jpg
https://i.imgur.com/Syx04CH.jpg

Excuse, the OC is applied in both the Core and the Mem, The PL is OK (not applied in both)

Unrelated issue.  I've got 3 rigs with 1060 3GB cards mining BEAM using GMiner.  Since the Equihash 150.5 algo requires just shy of 3GB of VRAM, I'm running those rigs on HIVEOS for now.

I'm monitoring then from AM as External miners by looking at port 42010.  This works fine except the earnings are not showing up.  When I right click and "Define coin for pool",  I select BEAM.  However, nothing is showing and if I go back to "Define coin for poo"l again, it has reverted back to BTC.

Also, on these External miners, when I click on the Pools tab, it shows the Grin29 pool my managed miners are connected to, not the BEAM pool these miners are actually mining on.  On my Managed miners, I can of course select the pool each rig is mining on, but for External rigs, this is not possible.  I do see the GPUs and hash rates as AM is picking that up from the GMiner API.

I don't know if something changes, but I seem to recall AM being able to show earnings for External miners in the past.

Yes, I only have a single host defined for this specific miner.  I did have a windows hosts defined previously that I deleted and replaced with the nvOC based one.  But this was also the case for the first rig that I converted from Windows to nvOC.  One note on this is that I had to restart AM before it would "pick up" the first rig that I switched from windows to nvOC.  And yes, I do have the "Linux" checkbox checked.


How do I access this log?

Also, will setting GPU clocks, power levels and fan control be a future option for the Linux remote agent?

btw, I really like the recently added native control of these with the windows version.  So much simpler than having to use the Afterburner remote agent.

One thing I found is that I had to re-create my GPU Clocking Profiles as the old ones did not include the temp target values.

Thanks for sharing all good points above.

I just wanted to give a short comment about why Awesome Miner don't have the algorithm names listed in the Managed Software dialog for Bminer.

For most mining software you have a concept of a command line argument specifying the algorithm, for example "phi2" that will give the result "-a phi2" for example. Using the Managed Software dialog you can customize this per algorithm to make it "phi2 -I 15" resulting in "-a phi2 -I 15".

For Bminer you don't have a separate command line argument for the algorithm name, instead it's part of the generated URL parameter that looks something like "cuckaroo29://username:[email protected]:13654" where Awesome Miner automatically set the part "cuckaroo29" based on the algorithm you have for the pool/coin. As the algorithm argument is part of the URL with Bminer, it's not possible to simply add additional command line arguments to it with the current design.

I agree with all troubleshooting steps above. The Diagnostics button is my best friend

Thanks for taking the time to try out this.

The Clocking Profile concept will be redesigned soon. The difficulty is that depending on which GPU you have, the clocking properties are different. Some cards support Target Temperature while others don't, and so on. By forcing you to setup a Clocking Profile based on what a specific card supports will make sure you don't try to set properties that isn't supported.

Please note that as soon as you save the Clocking Profile, it has no link to the specific miner or GPU, and it can be applied to any miner and and GPU. If you try to apply it to a GPU where a property isn't supported, that specific property will be ignored.

Just like before when you used MSI Afterburner, you have to define a Clocking Group if you have a mix of GPU types and want to specify a unique profile for each GPU type.

I've also updated the documentation for GPU overclocking with the Native feature and some troubleshooting suggestions:
https://support.awesomeminer.com/support/solutions/articles/35000086015-gpu-overclocking

The curve you are referring to, I assume it's the one in the "Voltage/Frequency curve editor" inside MSI Afterburner? Awesome Miner doesn't support this VFC feature yet but it's on the todo-list.

Thanks for your question

I assume you only have one single Host for this Linux miner defined in the Options dialog, Managed Hosts section. This host must have the "Linux" checkbox active and there shouldn't be any other hosts in the list connecting to the same computer. Can you verify this?

Can you also review the Awesome Miner application log file for any error message about this connection problem?

I know that a number of users have been running Remote Agent on HiveOS. I will however give nvOC a try to see if I can reproduce the problem you point out.

I've updated the documentation page for GPU overclocking. It also contains information about how to troubleshoot if needed.
https://support.awesomeminer.com/support/solutions/articles/35000086015-gpu-overclocking

Please note that for several of the OC settings, it's not possible to use MSI Afterburner to detect if it's applied correctly. For some settings, MSI Afterburner must be restarted for the new settings to show up and for a few other settings MSI Afterburner have it's own way of reading/setting clocking so any changes you define using other applications will not show up. For this reason I pointed out in the documentation that you should use the Reload-button in Awesome Miner to verify the settings.

It's no problem to leave MSI Afterburner installed. You can switch overclocking provider in the Options dialog as many times as you want.

Thanks for the suggestions.

About the notes field -  I understand your point here, but it's a bit similar to the discussion we had about Pool Profit Factor where a pool property would be related to a coin. I think this will have to be part of a concept later on where you have a Pools tab in addition to a Coins tab.

Default AM BMiner URL for version v6.1.12 is:

This requires CUDA 10 and I have not used this version since not all my rigs are migrated to the CUDA 10 drivers.

BMiner doesn't require the algo input parameter and thus the screen says you're unable to configure command line arguments in the said screenshot. I beleive it just negotiates and choose the algo automatically when connecting to pool stratums.

You may need to deliberately run diagnostics and see the output as to why it hasn't been working properly for you, either manually execute the command line through command prompt window, via .bat file, or if you want to do it in AM for a profit miner, probably easiest to do is to

-> Duplicate a profile (or just create a blank profile) for the problematic miner
-> Disable all software except bminer in profit profile
-> Optional (artificially inflate hashrate one of the BMiner supported algos that you have identified as problematic)
-> Assign the problematic miner with this temporary profit profile
-> Right click -> Diagnostics
-> Share the output of the diagnostics and see if anyone can pickup what might be wrong if it's not trivial.

https://i.imgur.com/x6VJAQH.jpg

https://i.imgur.com/f52Bn8k.jpg

https://bitcointalksearch.org/topic/annbminer-a-fast-equihashethashcuckaroo29z-miner-for-amdnvidia-gpus-1649-2519271
It can not be that, several people in my telegram group have the same url, which is the default one in AM and to them if it works, although it does not include the ticket of each a-l-g-o

It is the default download on your page, you can see it on the Ann link of BTC talk
and also I repeat that it is the one that AM includes by default

What is your Bminer url in the options> managed software> BMINER panel?

Upload manualy, dont be lazy

bminer-v14.3.0-cbb8683-amd64

@patrike or @mopiddo I have a problem. For days I have noticed that Bminer does not work well for me and I use other miners.
Today reviewing optiona> managed software when entering Bminer, I see that it has some enabled by default, but they do not have the text of "A-l-g-o" does not put Cucka29, or tensory. But he does not let me put it manually either.

If I enter any other miner, I see that if they have defined the "A-l-g-o" or I can define it manually, but it does not leave me.

How can I fix this, right now I can not use Bminer in anything.

Hello Patrike.

I think the Native overclocking does not work correctly or I can not configure it correctly. First, I do not understand why when we create an OC profile, we have to choose a particular miner and then a card. I guess a link is created between the OC profile and the miner or the card, right? I think this fact can be a headache because the ideal is to use it automatically in each algorithm that is needed applied only to profite profiles. Well, I'll explain. I have two different cards, one 1070ti and one 2070, and I have created several different OCs. The fact is that I apply the OC that I consider in the algorithms that I consider in each different profite profile, which is how we finally identify each one of the miners that we have. The fact is that when applied in a profite profile, the power limit is respected, but both the Core and the Mem are activated in the other profit profiles although they are not activated in the second. Logically, I keep the AB to see how the OC behaves. Can it be by using the option to clone profiles? Something does not work correctly  

To provide more information:

I started to realize when I noticed that some miners gave me an error. This happened because I have an OC for 2070 wild memory (+850) that the 1070ti can not support. The fact is that when the OC was activated for 2070, the 1070ti gave error by hanging the miner.

In this rig I only have two cards, but I have one rig with 6 cards and with 3 different types and I was going crazy

Another thing:

At the beginning I tried using the AB OC, since I have OC profiles in curve mode for the Core, that is, the AB allows you to create curved profiles for the different memory sets, so if your card is stable to +500 mem with Core +200 in power 70%, it is possible that with power 100% it is not, and these curved profiles allow you a stable OC setting only Mem, and vary the Core according to the Power that you set. The fact is that in the first tests, I left without clicking the Core, so that AM respected the core of the AB. The result is that when activated in a defined algorithm, it is respected, but at the same time the value is fixed in the AM, so in successive Oc profiles in AM, the core is fixed in advance, not respecting the changing value of the OC curve. I do not know if I explain myself very well.

This last one is the reason why I have rejected to do OC with the AB and to put myself with the native one, finding the problem exposed to me first.

I've been playing around with the Linux remote agent.  My first rig came up and started working right away, but I'm having issues with my 2nd rig.  When I look at the remote agent log, I see a bunch of these:


After a while it changes to:


So in AM, the miner shows up as "Service offline", but when I go to Managed Hosts and click "Test connection", I get the following message:

Successfully connected to Awesome Miner Remote Agent/Service 6.1.11

I uninstalled and reinstalled the remote agent, but I still have the issue.

I'm running nvOC 19.30 (Ubuntu 18.04), btw.  I like this distro since it is all setup for mining out of the box and has a bash file where I set the GPU clocks and power.

If there's another distro that provides basics I need, I'm not opposed to give that a go.  What Linux distro do you guys use with AM?

EDIT: Did a clean install of Ubuntu 18.04 LTS Server and after getting it all setup with the latest drivers, etc, I installed the remote agent.  Getting the exact same errors as on the nvOC distro.

I just did a test in a small rig. I have stopped MSI SERVER and in AM I have put the OC to be awesome miner.

I have forced several changes of something with different OC but in the rig the OC has not moved. I have left Aftherburner installed to know if the OC changes.

When I finished testing I went back to the usual with MSI SERVER and changed to AM and left it as usual.

That rig gave me problems, I had to go to the GPU and do the Mapping of the GPUs again to make it work.

As you said no info in your documentation please post the link here when you update it. I would like to use the most stable system. Sometimes the MSI SERVER stops working and you have to change the name of the folder in all the RIGS.

Right now I'm still with Aftherburne and MSI SERVER until there's more information on how to use your new OC version correctly.