Topic: [ANN] dstm's ZCash / Equihash Nvidia Miner v0.6.2 (Linux / Windows) - page 114. (Read 224961 times)

I dont understand how you can have 8 hours downtime ( from the miner point of view )
When this network failed happened i got notification from flypool and from my watchdog, i investigated, took me 10 minutes to switch back all my rigs on ewbf and to change the server from europe to asia.
In that day i made more coins then usual because a lot of miners didnt react to flypool europe server failure and the difficulty dropped.
The next day i switched back to dstm's miner.

If you mine and want uptime USE WATCHDOGS  SCRIPTS dont blame this on the mining software.

One from my rings on EWBF and he is work properly on this poll, without reconnect 2 hours in the time while dstm have a lot of reconnect.
I'm sure it was in the same time.

I'm able to reproduce this.
The server sends a TCP FIN packet i.e. closes the connection periodically.
Are you sure about your tests with other mining software?

Problem while mining ZEN ON  https://zenmine.pro

eu.zenmine.pro:9009

Every 2 minutes conection closed by server r:0

Could you check please?

EWBS work on this server without problem, with dstm every ~2-4 minutes conection closed by server r:0

I think it's fair to remove fee until a solution is found.

No, there was nothing rude in your comment, it's a perfectly valid position. It's very interesting how people are different in their judgments with respect to this.

Concerning ZM's flypool 'lock'. As far as I understand flypool has a high uptime something about 99.9x/8x%, this is what my tests during the last months suggest it also matches the experience people have made so far. ZM's implementation was based on this numbers.

The downtime because of ISP issues is really a stupid thing - zm clearly needs an improvement here. I don't really like the failover pool solution - it has some corner cases. I prefer much more to ignore the submission of dev fee shares in cases of server outages. This is less complex and perfectly reliable for you, however this easy to abuse and I'm not able to predict how many installations would do so.

That's why only 1/3 of my miners runs ZM miner, the rest are with EWBF and others.

yes I am greedy, i like money, i also like technology, and mining is the best of both worlds, but make no mistake, we ALL are here to make money, so greed or no greed, we all need to get even, to achieve ROI, hardware is expensive!, power is also VERY expensive!, i invest heavily in mining hardware, and i expect it to perform accordingly.
I setup everything to insane perfection, so when something that it's out of my hand, happens, and my miners stays offline for so long, i get mad!, not to dstm, he is working hard to make the miner great, he'll eventually get it right, but to the situation, yes I rant, so he fix it  can't help it  (sorry dstm if i was rude btw) LOL

I run linux, have my own custom distro, totally automated, with near 99.99% uptime.
testing something for 1 month in crypto could be suicide, here things change so rapidly that usually testing is the week-range, IMHO, not more!.

BTW: i haven't deployed ZM overnight, dstm knows my case, i tested it for 2 weeks on a 1080Ti, then decide to move to three rigs (6x1080) and then some more, stability was pretty good until this happens, and yes eu pools got fucked up by some major ISP nothing we can do about it, i inmediately changed to another location and another pool and then i realized that ZM was "locked" to eu port on fly pool, so i switched to EWBF (yes 8 hours LATE because i was dealing with a F* mess in a plant with some PLC controller , and had no time to check my miners, so yes i cursed at me as well for not taking 30 seconds to check my miners remotely)

anyway.....shit happens...
all online again
cheers
indkt.

Comments:

1) To these miners: it's not DSTM's fault that you lost money, it's your fault. Your greed got you in trouble.:

a) What business expects operational perfection in forecasting profit? So if you owned a trucking business, all the trucks would work perfectly? What would happen if a squirrel shorted out a breaker at the local power distribution station? There are some particular power distribution transformers that could take a week+ to replace because they are not stock items. Where I live, count on small animals, car crashes and lighting power outages.

b) You are stupid and greedy if you immediately deployed DSTM's miner across all your miners and then dumped on DSTM for any flaws. What was wrong with just running it on one mining setup? Even those bastions of software excellence, Microsoft and AMD, deploy updates that have problems/flaws. Why are you newbies dumping on one guy?

You guys need to come up with some sort of procedure for software deployment given all the noise you are making about how serious you are.

My procedure(not for everyone).

I'm conservative. I typically run at least 1 to 2 months behind any improved mining software. If the improvement is only a few %, I won't bite because crash free operation is worth a few per cent. If it's a few % hash increase plus a few per cent decrease in power consumption, this gets on my radar. I will wait several more weeks until people on the board I trust vouch for it. I then run this "new" software on one miner. Every crash adds a day of evaluation after a "corrective" homemade or goggle tweak. I have no expertise in determining whether there is bad stuff hidden in any software, this is my major concern. Yet another reason to treat this board's reliable software coders with respect and stop whining about their cut.

To me i get way more stable and better hash rate with Your Miner then ewbf. Your Fee is justified. Thank you!

Great! thank you for taking it serious, and worry about it, i'll wait until you sort this out and go back to your miner then.
please let us know when you got it
thank you
indkt.

That's perfectly valid ofc, no doubt - sry for the downtime.
I'm currently in contact with different pool owners, to be able to distribute pools across different networks.
I'm still not really happy about the failover draft, it has some corner cases that I don't like - like local network issues - more difficulties for people which have to tunnel all their traffic through a proxy - but I'm working on it.

dstm you never get credit for the work you do. I thank you for everything you have given me at least. There was some downtime but everyone has downtime somewhere.

Same here, 8 Hours offline with 30 x 1080 GPUs because of this problem.
not anymore.
I'll go back to EWBF
it connected instantly, 0 problems.

I lost a lot of money today.
sorry dstm but i can't afford problems like this, my profit marging is too thin right now to be offfline 8 hours for such a stupid issue.

DSTM - Thank you for the work you have put into this.  I know not everyone is appreciative right now, but I am.

All right! I don't speak for everybody, but to me 2 days sound cool.
I just had the impression that you didn't care and were about to do nothing about it.

Sry, I'm really trying to understand your intention, it looks like we are reading different things.

I've stated it clearly: reliability is of high priority - I'm working to handle such kind of situations reliably.
I don't want to provide an unreliable untested solution - so it will take a couple of days.
If this is not fast enough people have the option to use a different mining software, if there was no other mining software available providing an untested solution would be an option.

oh these trolls with chromosome disorder on this board. If you are so awesome, why don't you write a miner that's perfect ? Oh wait, missing half a brain and all you are capable of is drooling.

Very constructive.
You're actually the one who should shut it.

dude, we fking got it, you're salty. Run back to EWBF's and shut it, no one is forcing you to use this miner.

This was not a server outage but a network outage. That can happen everyday, maybe not that scale but it can happen anytime.
There is no hack to do, the real hack is your fee. You remove that and everything work fine.

The solution is simple: no connection, you drop the shares... nobody wins, but at least the miner don't crash. Is that a hack?

For now, you just make it look bad. It is unreliable because how your fee behave, and your reaction.
And telling people here on this forum, who has worked with you (yea, beta test), if your're not happy go use another miner: that's not very professional.

Lot's of us here had stability problems... well, it's becoming clear now. We have at least one explanation.
Not only we are paying fees, but it's turning against us.
But you don't wanna change anything, I get it.

I'm complaining, but I don't care, really. I just have a small rig. But think of those who have farms.
Some have spoken, you just ignore them.