Topic: [ANN] cpuminer-opt v3.14.2, open source optimized multi-algo CPU miner - page 6. (Read 10284 times)

Hi JayDDee,

Any chance you can re-introduce scryptjane  into your next release.
I'm stuck using 3.11.0 (scryptjane:20), and would love to see if the newer optimizations make a difference on Zen3. The extra L3 cache seems to make a small diff in all the newer miners.

Thanks
J

It will be in the next release, I just don't know when. In the meantime I posted a note in the Zergpool thread
and opened a git issue to document the workaround.

that worked for me, thanks.  If you can add parsing for stratum+ssl that'll work. If you don't at least add an explaination to use tcps instead of ssl in the pool URL.  I don't know if other pools use "stratum+ssl" rather than "stratum+tcps" in their guides.

Thanks, I didn't scroll down far enough.

I tried it with stratum+tcps and it worked. I can add parsing for stratum+ssl as a cli convenience
but I have no releases planned at this time.

on their front page says they do.  

-o stratum+ssl://.mine.zergpool.com: -u [-p ]
though it might not be working and the pool operator doesn't know it.

you have a better chance of finding out on their discord.
https://discord.com/channels/407610160580657164/412146651185020938

a message from march 29, 2022 on the discord channel
!pinpin [zergpool.com] — 03/29/2022
Happy to announce that we are adding TLS/SSL support also for all mining ports. All regular mining ports have its sibling now, which starts with 1. As example for sha256, TLS/SSL mining port is 13333. Happy mining, and please report if some of ports does not work for you in TLS mode. @everyone

I don't think Zergpool supports it. I'm not aware of many pools that do.

Is stratum+ssl still not well supported?  Doesn't seem to work with zergpool

cpuminer-opt doesn't support RandomX or any variants, xmrig does.

Hi, could you support to add  XDAG ?
XDAG is the first mineable DAG.
It is CPU mining , RandomX Algo.

XDAG's WEB:
xdag.io

Github：
https://github.com/XDagger

BitcoinTalk：
https://bitcointalksearch.org/topic/ann-xdag-new-community-based-cryptocurrency-first-mineable-dag-2552368

Based on your question in the Pooler thread it seems you're using it correctly but the mining difficulty is
too high to find any shares. Choose an algorithm/coin that is more suitable for CPU mining.

You will find all the needed Info here on the Github page  : https://github.com/JayDDee/cpuminer-opt

Specialy here and read all things : https://github.com/JayDDee/cpuminer-opt#requirements
Supported Algos : https://github.com/JayDDee/cpuminer-opt#supported-algorithms
But you should be also find all this Information on the first post in this Thread here.

Hello, is there any documentation on how to use the tool? Which coins and pools are supported? Tried with binance but no success

One of my goals was to make the miner somewhat modular to make it easier to add (or remove) algos.
Unfortunately it doesn't work with most of the newer miners/algos that use cmake & c++.
Also with most new coins choosing new unique algos the work to support new coins has increased.

As a result those new coins that can use the cpuminer mining architecture fork the miner to build their own.
My only disappointment with that approach is those who take open source code, close it, then charge a fee.

Thanks for the kind words.

Oh my lord....I must have had way too many txt documents up all at once, because I ended up successfully creating the cpu miner that I think is Wyvern or whatever Raptoreum has up on their website.  I had the READ ME of cpuminer-opt up and I was trying to figure that one out as well, so obviously my brain is a wee bit addled.  I just took a look and it does say right there in the READ ME doc that the gr algo isn't supported anymore.

The miner that's the object of this thread (cpuminer-opt) looks much more interesting now that I've begun to understand the workings of it.  I'll let the RTM miner run a little longer then try the other.  Thanks for catching my boo-boo.  I still give mucho props to JayDDee; that wasn't a mistake.

except this doesn't do RTM ,i.e. raptoreum , i.e. ghosterider.

you might have one of the forks from this one, where they did raptoreum.  there are a lot of forks out there where someone forked cpuminer to make enhancements on one algo.

I just stumbled across your cpuminer-opt program today and incidentally when I was Googling it I found this thread.  Just want to say it's nice to see the developer keeping his thread updated and on the straight and narrow.

Over the last year I've gotten interested in computers to the point of obsession, and that was after years of being into cryptocurrency.  That led me to learn about CPUs, which in turn led me to CPU mining--hence the reason why I'm writing this.  I downloaded cpuminer-opt and started running it on a mini PC that I have in another room, and right now it's tuning.  We'll see how much RTM I can mine.  It's not exactly my favorite CPU-minable coin but I'm having fun and it's certainly a learning experience.  Thanks for that!

cpuminer-opt-3.19.5 includes an enhanced stratum-keepalive feature to prevent lost shares.

TLDR stop here.

There really is no actual stratum-keepalive, it's a hack as described below. It doesn't really keep the
connection alive but it prevents the negative effects of the pool server timing out if no shares are
submitted for a long time.

CPU miners are challenged by pools that set a minimum difficulty too high for CPUs to find shares
regularly. After around 5 minutes with no shares submitted the pool will drop the connection but not
notify the miner. Often the miner detects the dropped connection only when submitting a share.
This results in a lost share and is noted in the summary log as stratum errors and share count mismatches.
Every mismatch in the counter represents a lost share. When shares are so rare this is a big loss.

The first stratum-keepalive was a futile attempt to send a harmless message to the server in the hope it would
prevent the timeout. It was also convenient that the protocol defines a specific method "mining.suggest-difficulty"
which perfectly suits the situation. Unfortunately this was not successful. Pools not only ignore the request for
lower difficulty it doesn't reset the's pool timeout either. In effect it's useless.

However, it will remain in case pools decide to support the method some time in the future.

The enhanced stratum-keepalive goes one step further, it will pre-emptively reset the stratum connection before
the pool times out and closes it. This will prevent the miner from sending a share over a dead connection. It also
eliminates wasted time hashing with a dead connection even if no share is found.

The suggest difficulty method is sent after 3 minutes with no shares submitted and the pre-emptive reset is done
after 4 minutes with still no shares submitted. Timing is not precise but the reset is usually done before 5 minutes.

The option is disabled by default and can be enabled with --stratum-keepalive. There should be no harm in using
even if timeouts aren't a problem.

Here's a sample of it in action:

Check out my new open source nVidia GPU miner with X16RT.

https://bitcointalksearch.org/topic/ann-suprminer-jdd-open-source-nvidia-gpu-miner-with-x16rt-5383533

Back to you, thanks for the great help.

Sounds good enough 
Have a great weekend!